<%BANNER%>

Design and Implementation of DDISMedia

Permanent Link: http://ufdc.ufl.edu/UFE0022865/00001

Material Information

Title: Design and Implementation of DDISMedia An Online Digital Media Library for Pest Diagnosis
Physical Description: 1 online resource (56 p.)
Language: english
Creator: Zhang, Shiwei
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: application, diagnosis, library, media, pest, web
Agricultural and Biological Engineering -- Dissertations, Academic -- UF
Genre: Agricultural and Biological Engineering thesis, M.E.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: A Distance Diagnostic and Identification System (DDIS) was developed at the University of Florida to assist in pest diagnosis. Since its original deployment in 1999, thousands of digital pest photos have been collected and catalogued in DDIS. Our study involves DDISMedia, a user-friendly rich Internet digital media library application developed to assist specialists in pest diagnosis. DDISMedia, coupled with the DDIS diagnosis environment, provides a new tool for specialists to improve diagnosis and identification of pests. The DDISMedia has a peer-reviewed media database, which contains a collection of digital media produced by specialists of different disciplines including plants, plant insects and diseases. This system allows registered users to submit pest images or video clips into the database. Techniques used in implementing DDISMedia include a peer review process, library management, and extensive search functions. The media library of the DDISMedia could also be used for research, teaching/learning, and educational programs.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Shiwei Zhang.
Thesis: Thesis (M.E.)--University of Florida, 2008.
Local: Adviser: Zazueta, Fedro S.
Local: Co-adviser: Momol, Timur M.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0022865:00001

Permanent Link: http://ufdc.ufl.edu/UFE0022865/00001

Material Information

Title: Design and Implementation of DDISMedia An Online Digital Media Library for Pest Diagnosis
Physical Description: 1 online resource (56 p.)
Language: english
Creator: Zhang, Shiwei
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: application, diagnosis, library, media, pest, web
Agricultural and Biological Engineering -- Dissertations, Academic -- UF
Genre: Agricultural and Biological Engineering thesis, M.E.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: A Distance Diagnostic and Identification System (DDIS) was developed at the University of Florida to assist in pest diagnosis. Since its original deployment in 1999, thousands of digital pest photos have been collected and catalogued in DDIS. Our study involves DDISMedia, a user-friendly rich Internet digital media library application developed to assist specialists in pest diagnosis. DDISMedia, coupled with the DDIS diagnosis environment, provides a new tool for specialists to improve diagnosis and identification of pests. The DDISMedia has a peer-reviewed media database, which contains a collection of digital media produced by specialists of different disciplines including plants, plant insects and diseases. This system allows registered users to submit pest images or video clips into the database. Techniques used in implementing DDISMedia include a peer review process, library management, and extensive search functions. The media library of the DDISMedia could also be used for research, teaching/learning, and educational programs.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Shiwei Zhang.
Thesis: Thesis (M.E.)--University of Florida, 2008.
Local: Adviser: Zazueta, Fedro S.
Local: Co-adviser: Momol, Timur M.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0022865:00001


This item has the following downloads:


Full Text

PAGE 1

1 DESIGN AND IMPLEMENTATION OF DDISM EDIA: AN ONLINE DIGITAL MEDIA LIBRARY FOR PEST DIAGNOSIS By SHIWEI ZHANG A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ENGINEERING UNIVERSITY OF FLORIDA 2008

PAGE 2

2 2008 Shiwei Zhang

PAGE 3

3 To my family

PAGE 4

4 ACKNOWLEDGEMENT I expres s my sincerest gratitude to the members of my supervisory committee, Dr. Fedro Zazueta, Dr. Tim Momol, Dr. Ji annong Xin, and Dr. Thomas Burks, for their valuable advice, guidance, and support in the past three years. I thank Dr. Zazueta, my advisor, for teaching me research skills, and all his efforts to solve my problems promptly in achieving my masters degree. I thank Dr. Momol for supporting me my research assistantship, and his valuable fee dbacks and comments for this project. I am also grateful to Dr. Xin for his great input in my work. He taught me new information technologies and guided me throughout my work. It was lucky for me to work with Dr. Xin for three years. I also thank Dr. Burks for his understanding and support. Special thanks go to my parents for their tr emendous love and support in my life. Finally, my heartfelt appreciation goes to my wife, Na n Feng, for her love, help, encouragement, and accompaniment. Without them, I could not achieve this goal.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGEMENT...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........7 LIST OF FIGURES.........................................................................................................................8 ABSTRACT.....................................................................................................................................9 CHAPTER 1 INTRODUCTION..................................................................................................................10 1.1 Background.......................................................................................................................10 1.2 Existing Plant and Pest Digital Libraries.......................................................................... 11 1.2.1 Forestry Images...................................................................................................... 11 1.2.2 PDIS Library..........................................................................................................12 2 SYSTEM OVERVIEW..........................................................................................................13 2.1 Requirement Analysis....................................................................................................... 13 2.2 System Design.............................................................................................................. ....14 2.3 Technologies and Tools.................................................................................................... 16 2.3.1 HTML.....................................................................................................................16 2.3.2 JavaScript...............................................................................................................17 2.3.3 Java Servlet.............................................................................................................18 2.3.4 JSP..........................................................................................................................18 2.3.5 Apache FileUpload.................................................................................................19 2.3.6 SQL.........................................................................................................................19 2.3.7 JDBC......................................................................................................................19 2.3.8 Ajax........................................................................................................................20 3 SYSTEM IMPLEMANTATION........................................................................................... 22 3.1 Media Submit............................................................................................................... .....22 3.1.1 Data Schema...........................................................................................................22 3.1.2 Work Flow.............................................................................................................. 24 3.2 Media Review............................................................................................................... ....25 3.2.1 User Role................................................................................................................26 3.2.2 Data Schema...........................................................................................................26 3.2.3 Work Flow.............................................................................................................. 28 3.3 Media Search/View.......................................................................................................... 29 3.3.1 Media Search: Extens ive Search Options...............................................................30 3.3.2 Media Browser....................................................................................................... 30

PAGE 6

6 4 USER INTERFACE DESIGN...............................................................................................34 4.1 Media Submission UI....................................................................................................... 34 4.2 Input Suggestions fo r Pest/Host Names........................................................................... 36 4.3 Media Review UI..............................................................................................................37 4.3.1 Working List UI..................................................................................................... 37 4.3.2 Assignment UI........................................................................................................37 4.3.3 Review UI............................................................................................................... 38 4.3.4 View Comments UI................................................................................................ 39 4.3.5 Decision UI.............................................................................................................39 4.4 Media Search/View UI.....................................................................................................40 4.4.1 Media Search UI..................................................................................................... 40 4.4.2 Media View UI....................................................................................................... 40 4.5 Interfaces within DDIS.....................................................................................................42 5 SUMMARY, CONTRIBUTIONS, AND FUTURE DIRECTIONS..................................... 51 5.1 Summary...........................................................................................................................51 5.2 Contributions....................................................................................................................51 5.3 Future Directions.......................................................................................................... ....52 LIST OF REFERENCES...............................................................................................................54 BIOGRAPHICAL SKETCH.........................................................................................................56

PAGE 7

7 LIST OF TABLES Table page 3-1 Data schema of Table img_lib........................................................................................ 31 3-2 Data schema of Table img_media.................................................................................. 31 3-3 Data schema of Table img_editor_process.................................................................... 32 3-4 Data schema of Table img_reviewer_process................................................................ 32

PAGE 8

8 LIST OF FIGURES Figure page 2-1 Integral structure of DDISMedia....................................................................................... 21 3-1 The ID generation flow chart in Media Submit................................................................. 33 3-2 Peer review process in DDISMedia................................................................................... 33 4-1 Sample information form................................................................................................... 43 4-2 Location form.............................................................................................................. .......43 4-3 Geo-code input panel....................................................................................................... ..43 4-4 Additional description form and copyright notes f orm...................................................... 44 4-5 Photographer/video producers form.................................................................................. 44 4-6 Upload Media form............................................................................................................44 4-7 Example of Ajax-based input suggestions......................................................................... 45 4-8 Example of In Review working list UI for editor-in-chief/editor ..................................45 4-9 Example of Assignment UI for editor-in-chief.................................................................. 45 4-10 Example of Review UI for reviewer.................................................................................. 46 4-11 Example of user interface when selec ting Accep t with revision in Review UI............. 46 4-12 Example of View Comm ents UI for reviewer...................................................................47 4-13 Example of Decision UI for editor-in-chief/editor ............................................................47 4-14 Media Search UI................................................................................................................47 4-15 Example of Tag Cloud UI.................................................................................................. 48 4-16 Advanced Search UI..........................................................................................................48 4-17 Example of square image viewer....................................................................................... 49 4-18 Example of thum bnail im age viewer................................................................................. 49 4-19 Media viewer.............................................................................................................. .......50 4-20 DDISMedia embedded in DDIS pest diagnosis................................................................. 50

PAGE 9

9 Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Engineering DESIGN AND IMPLEMENTATION OF DDISM EDIA: AN ONLINE DIGITAL MEDIA LIBRARY FOR PEST DIAGNOSIS By Shiwei Zhang December 2008 Chair: Fedro Zazueta Cochair: Tim Momol Major: Agricultural and Biological Engineering A Distance Diagnostic and Identification System (DDIS) was developed at the University of Florida to assist in pest di agnosis. Since its original deploy ment in 1999, thousands of digital pest photos have been collected and catalogued in DDIS. Our st udy involves DDISMedia, a userfriendly rich Internet digital media library applic ation developed to assist specialists in pest diagnosis. DDISMedia, coupled with the DDIS di agnosis environment, provides a new tool for specialists to improve diagnosis and identificatio n of pests. The DDISMedia has a peer-reviewed media database, which contains a collection of di gital media produced by sp ecialists of different disciplines including plants, plant insects and dise ases. This system allows registered users to submit pest images or video clips into the database. Techniques used in implementing DDISMedia include a peer review process, libr ary management, and extensive search functions. The media library of the DDISMedia could also be used for research, teaching/learning, and educational programs.

PAGE 10

10 CHAPTER 1 INTRODUCTION 1.1 Background The web-based Distance Diagnostic and Identi fication System (DDIS) was developed to provide a unique environm ent to conduct pest di agnosis through the Inte rnet. Since its first release in 1999, DDIS has served as an effec tive tool to assist the collaboration and communication among first detectors, extensi on specialists and diagnosticians to share information on plant insects and diseases. This system uses field data and digital media as primary sources to facilitate the diagnosis and identification of plant disease, insect, weed, invasive species, plant management, and livestock (Xin et al., 1998; 2003). The primary information transferred from agri cultural extension agen ts to diagnosticians contains massive photographic imag es. Therefore, thousands of these photographic images have been collected since 1999, serving as an archived database for research, educational programs, and teaching/learning purpose in diagnosing plant diseases (Holmes et al., 2000). However, the following defects in the current system interfer e with the management and usability of the photographic images. First of all, it lacks effective image manage ment component. For instance, inconsistency of image quality cau sed by the fact that the major ity of the photos were taken by extension agents, not a professi onal photographer in laboratories of agricultural fields, usually damages further application of DDIS in the fiel d of research and educ ational programs. The second defect derives from the fact that this syst em does not have the capability of direct image searching. In DDIS, images are by-products of DDIS samples, which means users can not obtain images unless they browse a DDIS sample. In addition, DDIS samples are only available to approved registered DDIS users. Agricultural exte nsion clientele and othe r external users who are interested in getting information from DDIS may not be able to access the system.

PAGE 11

11 Because of defects of the current DDIS disc ussed above and the feedbacks from system managers and users, it is desi rable to develop a digital media library in DDIS to serve as an important supplement of DDIS. This media library would include high quality video clips and/or photos of pests, plants, and invasive species that are submitted from any registered users worldwide. To ensure the media are described correctly and completely by their corresponding text information, an accurate and expertise-base d assessment process has to be performed. Once media are approved after the assessment, their can be retrieved and browsed by the public. In addition, a powerful direct search ing functionality is also a nece ssity to help users obtain their inquiries efficiently. To solve the problems di scussed above, an online digital media library: DDISMedia is designed and developed in this research. DDISMedia will be embedded into DDIS and eventually serve as an importa nt resource for DDIS in pest diagnosis. 1.2 Existing Plant and Pest Digital Libraries Bef ore the design of the DDISMedia, other existing online digital image libraries were visited and explored. Forestry Images and PDIS Library are two represen tatives of the existing digital image libraries, and some f eatures of these two libraries ar e worth using for references in the design of DDISMedia. 1.2.1 Forestry Images Forestry Images provid es an ar chive of high quality images re lated to forest health and silviculture, with particular em phasis on educational applications. In most cases, images shown in this system were taken by outside photographers. This system mainly collects images from forest pests, trees and plants, silviculture, invasive species, and wild life. Currently, this system contains approximately 90,000 images in about 12,000 subjects. Moreover, one feature of this system is to allow public users to search images based on different options, such as location, subject, and photographer. However,

PAGE 12

12 invariably errors may occur during the image su bmission process due to the fact that only the submitter has the responsibility to ensure the te xt information describing the image is accurate. In other words, the Forestry Images system lacks quality control procedure for the submitted images. 1.2.2 PDIS Library PDIS Library is designed to provide a large database of pl ant diagnostic related im ages that faculty, students, and government agencies can use for publications, presentations, and reports. This image library is searchable by: taxonomic serial number, common name synonyms, scientific name synonyms, disease name, pest name, photographer, submitting institution, county and state, description, and image size. In this system, each image is approved by specialists and classified using taxonomic and common name synonyms. Only individuals in the PDIS participating institution can submit images into the library. Like Forestry Images, PDIS Library does not support video clips ei ther, and only allows single image to be uploaded in one submission.

PAGE 13

13 CHAPTER 2 SYSTEM OVERVIEW 2.1 Requirement Analysis Before the design of DDISMedia, potential feat ures and requirem ents for this system have been discussed and determined. First of all, as a web appli cation in the area of agricult ure and biology, DDISMedia would likely face users without much computer and in ternet knowledge. Thus, a well-designed graphic user interface (GUI) is one of the key factors in the system design. The following three aspects are considered in the design of the GUI, (1). Ea ch page should emphasize one topic to minimize possible distraction and confusi on for users, (2). Operation fl ow of the system should be straightforward and navigate users to their destinations easily, (3 ). Online Help with conceptual or background information will be added to the user interface. Secondly, the media submission process of th e DDISMedia should be user-friendly and reliable. On one hand, the submission interface s hould be simple to use. Complicated user interface may confuse and discourage people from using the system. Moreover, the DDISMedia should be designed to minimize user input requi red for the submission process. On the other hand, a reliable file upload module should be utilized in the DDISMedia. As discussed in Chapter 1, large size files, such as video clips and photographic images, are major components in user submission; therefore, a reliable upload mo dule will likely shorten the submission time and minimize the failure possibility of submission. Thirdly, it is also necessary to implement a method to ensure the high quality of media in DDISMedia. High quality means high-resolution of electronic media, and accurate and complete text information corresponds to the media, such as sample common name, scientific name, and family. The DDISMedia is designed as an importa nt supplement and references for DDIS in pest

PAGE 14

14 diagnosis, and will also be used as a reliable resource for research and educational programs in the future, thus, high quality of the media is a key factor for DDISMedia to success. Chapter 3 will discuss the methodology of media quality control. The ability to perform extensive search is th e fourth requirement fo r DDISMedia, and the following criteria are considered, (1). The search engi ne should be able to re trieve data in a short time and have the intelligence to give a result set highly related to users inquiries, (2). Given the same search intelligence, the user interface for sear ching should be easy to use, (3). The display of search images should be well-organized and pr ovide user concise text information. The search function of DDISMedia will be open to public; th erefore, its quality w ill greatly affect the popularity of DDISMedia. Last but not the least, since DDISMedia will eventually serve as an important resource for DDIS in pest diagnosis, a user interface should be implemented in DDIS diagnosis page to direct diagnosticians into DDISMedia to perform related search. 2.2 System Design This section is an overview of the design of DDISMedia based on the requirem ents stated in the previous section. Figure 2-1 gives an integral structure of DDISMedia, which includes four major components: (1) Media Submit, (2) Media Review (3) Media Search/View, and (4) Assisted Diagnosis. Any registered user can submit into DDISMed ia via Media Submit. The registration is open to the general public. Users in DDIS, IPDN (International Pest Diagnosis Network) and CRDN (Caribbean Regional Diagnos is Network) are automatically granted submission right into DDISMedia. File uploading, as a key point in Media Submit, uses the Apache FileUpload package (Section 2.3.5) to ensure reliability and stability. Although submitters are allowed to

PAGE 15

15 decide at will the number of media to be uploa ded per time, it is strongly recommended to limit the total size of one uploading to 50 MB in DDISMedia. Besides, most text fields, except personal information, are optional in this component in order to speed the submission process. All missing or incorrect informa tion will be rectified in the next component: Media Review. The Media Review component is used to ensu re the quality of media library. Each new submission must go through a peer review proc ess, where the editorin-chief, editors, and reviewers are responsible for the quality of media and also make sure the text information associated with the media is accurate and comple te before its acceptance. If an error is found in the text information, the editor-in-chief/edito r will make revision based on reviewers recommendations. The Media Review component also provides a rich user interface that allows the editor-in-chief, editors and reviewers to inte ract during the review pr ocess. Only accepted media are ready to be retrieved in th e following Media Search/View component. Another significant component in DDISMedia is Media Search/View, which is further divided into two parts: search module and browse module. The search module provides five search options: (1) search by keyw ord, (2) advanced search, (3) browse all, (4) associated search, and (5) tag-based search. Keyword search is a generic inexact-match s earch based upon userranked keywords. Advanced search allows user to select different terms to specify their search criteria. Browse all enables user browse square images of all accepted media in database. Associated search option enables user search related samples to current search result through genius, species, family, media submitter or phot ographer as associated fi elds. Tag-based search extracts and ranks popular tags in current database and allows user to search based on each tag. The browse module retrieves complete information of one submission and provides a userfriendly platform for user to browse me dia and its associated text information.

PAGE 16

16 The Assisted Diagnosis component connect s DDIS and DDISMedia, and provides DDIS diagnosticians a functionality to us e the digital media library as a reference tool during their pest diagnosis. Other components in DDISMedia include da tabase connection component, by which all other components communicate with database server and suggested fill component that provides users input suggestions according to their inputs. 2.3 Technologies and Tools To im plement functions and components disc ussed in the previous section, the following technologies and tools were studied and utilized in the development of DDISMedia, including: HTML, JavaScript, Java Servlet, JSP, Apache FileUpload package, JDBC, SQL, and Ajax. 2.3.1 HTML HTML (HyperText Markup Language) is the prim ary markup language for Web pages. It describes the structure of text-based informati on in a document by denoting certain text as links, headings, paragraphs, lists, and so on, and to s upplement that text with interactive forms, embedded images, and other objects. HTML is wr itten in the form of tags, surrounded by angle brackets. HTML can also describe, to some degr ee, the appearance and semantics of a document, and can include embedded scripting language code (such as JavaScript introduced below) which can affect the behavior of web browsers and other HTML processors (Musciano et al., 2007; Willard, 2007). HTML markup consists of several key component s, including elements and their attributes, character-based data types, and character refere nces and entity references. A simple example is given out below:

PAGE 17

17 DDISMedia

Welcome to DDISMedia!

These eight lines of HTML code displays text Welcome to DDISMedia! on the web browser within the page titled DDISMedia. HTML is the mostly used technology in the programming on client-side in DDISMedia. 2.3.2 JavaScript JavaScript is a scripting langua ge often used for client-side web developm ent. It is a dynamic, weakly typed, prototype-based language, it also supports constructing new functions during the execution of a program, st oring them in data structures, passing them as arguments to other functions, and returning them as the va lues of other functions (Goodman et al., 2004; Flanagan, 2006). The primary use of JavaScript is to write f unctions that are embedded in or included from HTML pages and interact with elements of the pa ge. Some simple examples of this usage are: Popping up a dialogue to notify user an error message. Validating web form input values to ensure th at they will be accepted before they are submitted to the server. Changing text size or color as the mouse cursor moves over them: this effect is often used to attract the user's attention to important contents on the page. JavaScript code can run locally in a user's browser and can respond to user actions quickly, making an application feel more responsive. JavaScript code can also det ect user actions which HTML alone cannot, such as individual keystrok es. Therefore, JavaScript is used in the

PAGE 18

18 development of DDISMedia to help interact be tween users and pages and reduce burden of the server. 2.3.3 Java Servlet A Servlet is an object that receives a re quest and generates a response based on that request. The basic servlet package defines objects representing serv let requests and responses, as well as ob jects reflecting the servlet's configura tion parameters and execution environment. It contains package defining Hypert ext Transfer Protocol (HTTP)-spe cific subclasses of the generic servlet elements, including session management objects that track multiple requests and responses between the Web server and a client. Java Servlet Application Program ming Interface (API) allows a software developer to add dynamic content to a Web server using Java pl atform. Java Servlet defines the expected interactions of a Web co ntainer and a servlet. A Web container is essentially the component of a Web server that interacts with the servlets. Th e Web container is responsible for managing the lifecycle of servlets, mapping a Uniform Resource Locator (URL) to a particular servlet and ensuring that the URL requester has the correct access rights (Hunter et al., 2001; Perry, 2004; Hurniawan, 2002). Usually Java Servlet is used in conjuncti on with JavaServer Pages (JSPs), which is introduced below. In the implementation of DDISM edia, such conjunction is widely used: Java Servlet API is used in server-side programming, wh ile JSP is used to pass parameters values between the server and clients. 2.3.4 JSP JSP is a Java technology that allows softwa re developers to dynam ically generate HTML or other types of documents in response to a Web client request. The technology allows Java code and certain pre-defined actions to be embedded into static content, such as HTML codes

PAGE 19

19 (Perry, 2004; Bergsten, 2003). In DDISMedia, this technology is used to pass values of parameters between the server and clients. 2.3.5 Apache FileUpload The Apache FileUpload package is an open-s ource and robust library handling file uploads and is developed and m aintained by the Apache Software Foundation. FileUpload creates new file items using a FileItemFactory, which gives File Upload most of its flexibility. The factory has ultimate control over how each item is created. The factory implementation that currently ships with FileUpload stores the item's data in memory or on disk, depending on the size of the item. Due to its reliability and stability in JSP/Java Servlet environment, Apache FileUpload package is selected as the file uploading tool for DDISMedia. 2.3.6 SQL Structured Query Language (SQL) is a data base com puter language designed for the retrieval and management of data in relati onal database management systems (RDBMS), database schema creation and modification, and database object access control management (Ramakrishnan et al., 2002; Groff, 2002). Being a standard interactive and programming language for querying and modifying data and managing databases, SQL is mainly used to form commands that allow the retrieval, insertion, updating, and deletion of data in DDISMedia. 2.3.7 JDBC Java Database Connectivity (JDBC) is an API for the Java programm ing language that defines how a client may access a database. It provides methods for querying and updating data in a database. JDBC is oriented towards relational databases. JDBC allows multiple implementations to ex ist and be used by the same application. JDBC provides a mechanism for dynamically loading the correct Java packages and registering

PAGE 20

20 them with the JDBC Driver Manager. The Driver Manager is used as a connection factory for creating JDBC connections. JDBC connections support creating and executing statements, which may be update statements such as SQL's CREATE, INSERT, UPDATE and DELETE, or query statements such as SELECT. Additionally, stored procedures may be invoked through a JDBC connection. JDBC API is used in DDISMedia to operate database connection. 2.3.8 Ajax Asynchronous JavaScript and XML (Ajax), is a group of interrelated web developm ent techniques used for creating interactive web applica tions or rich Internet applications. With Ajax, web applications can retrieve data from the server asynchronously in the background without interfering with the display and behavior of the existing page. In many cases, different pages on a website share common contents. Using traditional methods, these common contents would have to be reloaded on every request. However, using Ajax, the common contents do not have to be updated on every request, which consequently reduces bandwidth usage and load time. The use of Ajax also allows th e client's Web browser user interface to be more interact ive and to respond quickly to input s, and also allows sections of pages to be reloaded individually. However, with the application of Ajax, an obvi ous problem may occur. If a users browser does not support Ajax or JavaScript, or simply has JavaScript disabled, Ajax module will lose its functionality. In addition, the lack of a standard body behind Ajax leads to the absence of a best practice to test Ajax applic ations (Crane et al., 2005; Mahemoff, 2006; Paulson, 2005). In DDISMedia, Ajax is mainly used to provi de input suggestions and pop up online Help and enhance user rich Internet experience.

PAGE 21

21 Figure 2-1. Integral st ructure of DDISMedia

PAGE 22

22 CHAPTER 3 SYSTEM IMPLEMANTATION This chapter discusses the im plementation of three major components in DDISMedia, including Media Submit, Media Review, and Media Search/View. 3.1 Media Submit The Media Subm it component collects information for DDISMedia and save them into database. Media Submit is open to any registered users and registration is open to the public. Registered users in DDIS, IPDN, and CRDN ar e automatically granted submission rights to DDISMedia. Submitted information is usually divided into two parts: text information and digital samples. In one submission, users are allowed to provide one copy of text information and upload as many digital samples as they will. But the total size of one submission is strongly recommended to be limited to 50 MB. 3.1.1 Data Schema In the datab ase level, there are two tables storing submitted information. One table titled img_lib stores text information for each submission. Table 3-1 lists the database schema of this table. Attribute LibID is the library ID and each submission is assigned a unique library ID once the submission enters the database. This unique ID is currently composed of eight characters: two numeric characters representing submission year, one dash character, and five random numeric characters created by a Java program on server-side. Attributes CommonName, G enus, Species, and SubS pecies represent the common name, genus, species, and sub species of the sample in a submission, respectively. Genus, species, and sub species constitute the scientific name of the sample. Similarly, HostPlant,

PAGE 23

23 HostGenus, HostSpecies, and HostSubSpeci es represent common name and scientific name of a host. Attribute SampleType is a required field for each submission. Same as in DDIS, six sample types exist in DDISMedia, and they are: (1 ) plant disease, (2) insect (plant), (3) insect (non-plant), (4) plant/weed, (5) managemen t/physiology/nutrient, and (6) livestock. Attribute Family specifies the family to which a sample belongs. It might not be an applicable attribute since some sample types do not have a family. Attributes County, City, State, Count ry, Longitude, a nd Latitude are all geographical attributes which represent the location of samples. These attributes could be used in sample distribution map, which is a function in DDIS, and also for statistical purpose. Attribute CollectionDate is the date when the sample was collected, and is a required field in submission interface. Attribute AdditionalDscp allows submitter to add additional descriptions to the sample. Attributes PhotographerName, PhotographerInstitute, and PhotographerEmail save personal information of the author of digital samples. Attribute SubmitDate files the date when a sample is submitted. Attribute SubmitBy refers to the user ID of a submitter. This user ID is either from DDISMedia, DDIS, IPDN or CRDN. Attribute Status indicates the current status of a sample in the library. The default status value is meaning the submission is new and waiting for a peer review process that (Section 3.2). Status means the sample has been revi ewed and approved to be accepted. Status denotes the sample is in review. If a sample ha s a status it means the sample has been

PAGE 24

24 reviewed but rejected. Only samples with status are ready to be retrieved and browsed by public users. Attributes EditorUpdateDate, EditorBy, EditorDecision, and E ditorComments are four attributes used by Media Review component (Section 3.2). Attribute CopyrightNote allows submitter to claim copyright for their submissions. In Media Submit, table img_media is used to store information for each digital media. Table 3-2 lists the data schema of Table img_m edia. Each attribute shown in Table 3-2 is described in details in the following. Attribute LibID in Table 3-2 must be an ex isting library ID in Table img_lib. Text information in Table img_lib and media inform ation in Table img_media for one submission share the same library ID. Library ID in Table 3-2 is not necessary to be unique, since it is possible that there are several digital media in one submission. Attribute ImgID is a serial number assigne d to each submitted media. Combination of this attribute and LibID above is treated as the key of Table 3-2. Attribute FileName is the original name of the media file before being submitted to the database. Once a media is submitted to the database, it will be renamed based on the combination of LibID and ImgID. Attribute FileType indicates the file type of the submitted media file. Attribute Media stores binary data of submitt ed media file and serves as a backup in case any emergency occurs on the physical disk of media storing server. 3.1.2 Work Flow When a subm ission request is received by Servle ts, a Java program is called to generate a random string and this string is assigned to the submission as its unique ID, as introduced above. To ensure the uniqueness of th e library ID in Table img_lib validation is required. If the

PAGE 25

25 generated ID exists in the current database, the process of ID generation will be repeated until a new unique ID obtained. An ID generation flow chart in Media Submit is shown in Figure 3-1. Once a library ID is assigned to a submission, the Apache FileUpload package is used to obtain text information and media fi le list from HTML pages. By us ing this file upload package, media files are loaded into memory and are read y to be processed. At this point, Media Submit writes a serial number onto the image, generate s square copy, thumbnail copy, and normal-sized copy of the original image, and finally writes these images (i.e., square copy, thumbnail copy, normal-sized copy, and original copy) onto physical disk of media storing server. In terms of a video clip, except for the above steps, Media Subm it needs an additional step to capture a picture from the clip before image processing. After media upload is completed, the text information and media information is written into the database. Considering the possibility that one submission may contain more than one media file, text information and media informati on are stored in two different tables in the database. All text information describing digital samples with unique ID is stored in the Table img_lib, while media information is stored in the Table img_media. Once a new submission is input into the database, its status is assigned value which means the submission is new and waiting for peer review and approval. 3.2 Media Review DDISMedia is a specialized pest and plant m edia database, where expert knowledge is required to not only ensure quality of the media, but, more importantly, verify completeness and correctness of text information associated with the media. Media Review is such a component designed to conduct this task. A procedur e named peer review process is used.

PAGE 26

26 3.2.1 User Role Sim ilar to the journal publica tion review process, media subm itter, editor-in-chief, editor, and reviewer are participating in the peer review process. The media submitter submits new submissions to the editor-in-chief. The editor-in-chief takes in charge of assigni ng new submissions to editor or reviewer. A new submission always goes to the editor-in-chief first, who then makes arrangement according to the sample type of this submission. Editor-in-c hief has the right to make a final decision of accepting or rejecting a submission. Generally speak ing, only one editor-in-chief is assigned to the peer review process. The editor is responsible for assigning subm issions received from editor-in-chief to reviewer. Editor can also make a final decision of accepting or rejecting a submission. There are usually several editors within different specialties. The reviewer is a specialist in a certain ar ea and provides comments and recommendations to editor-in-chief and editor. The peer review process does not have a fixed reviewer group in DDISMedia. However, thanks to their profe ssional knowledge, all DDIS specialists can potentially be selected as reviewers in this process. 3.2.2 Data Schema In database level, three tables are used in the peer review process. Attributes EditorUpda teDate, EditorBy, EditorDecisio n, and EditorComments in Table img_lib (Section 3.1) are updated when a final decision is made. Table img_editor_process stores editors actio ns in the peer review process. Its data schema is shown in Table 3-3. Each attribute show n in Table 3-3 is described in details in the following. Attribute UserID represents ed itors user ID in DDIS system.

PAGE 27

27 Attribute LibID is the libra ry ID of submission in Table img_lib. Combination of LibID and UserID is treated as the key in this table. Attribute ReceiveDate records the date when the submission is assigned to the editor. Attribute Decision is an attribute repres enting a decision made by the editor. After a decision is made, one of the following three values is assigned to this attribute. Value indicates Accept; value means Accept with revision; value represents Reject. Attribute DecisionDate records the date when the decision is made. By default, the value of this field is NULL, which m eans editor has not made any decision. Attribute Status is an attribute which indicate s the status of a submi ssion in editors peer review process. Value means the submission is in editors Unassigned list and waiting for review and approval. Value represents th e submission is in editors In Review list. Attribute Comment stores messages from editor to submitter when they make a decision. Table img_reviewer_process is used to record comments and recommendations from reviewer. Table 3-4 shows the data schema of Table img_reviewer_process. Each attribute shown in Table 3-4 is describe d in detail in the following. Attribute AutoID is a unique ID representing a record in this table. It is automatically generated by the database system. Attribute LibID is a foreign key wh ich can be found in table img_lib. Attribute UserID is reviewers user ID in DDIS. Attributes from CommonName through Family in Table 3-4 have the same meaning as their counterparts in Table img_lib (Section 3.1).

PAGE 28

28 Attribute Comments_Private is the field wh ere reviewer inputs comments to editor-inchief/editor only. Attribute Com ments_Public is the field for reviewer to store comments for both submitter and editor-in-chief/editor. Attribute ReviewerDecision represents revi ewers decision on the submission. This field by default is set to be NULL, which indicate s reviewer has not conducted any review. One of the following three values is assigned when a d ecision is made by reviewer. Value represents acceptance of the submission, value means acceptance of the submission but revision is needed according to reviewers comments, valu e indicates rejection of the submission. Attributes AssignDate and ReviewDate r ecord the dates when the submission is assigned and reviewed, respectiv ely, for management purpose. 3.2.3 Work Flow The peer review process m imics a journal public ation review process. A technical editorial board including one editor-in-chief and several editors is formed first to manage the whole process. Figure 3-2 shows the peer review process in DDISMedia. As shown in Figure 3-2, a new submission is first included in the Unassigned list of editor-in-chief with a status value of The edit or-in-chief next assigns editors or reviewers to initiate a review process. Then the submission is placed into the editorin-chiefs In Review list by changing its status from to in Tabl e img_lib. At the same time, the access to the submission is authorized to the assignee. Ema ils are also sent out by the system to notify submitter and assignee the initiation of the peer review process. If the submission is assigned to an editor from the editor-in-chief, a new record containing the library ID and editors user ID is inserted in Table img_editor_process with a status value of which means the submission is curren tly in editors Unassigne d list. To assign a reviewer for the submission, a new record with the library ID and reviewers user ID is inserted

PAGE 29

29 in Table img_reviewer_proce ss, and the submission is then included in the editors In Review list. Emails are sent out to inform submitter, reviewer, and editor-in-chief that the submission has been assigned to reviewer. If the submission is assigned to a reviewer from the editor-in-chief or editor, a new record containing the library ID and revi ewers user ID is inserted in Table img_reviewer_process. The reviewer is able to see submissions to be re viewed in his/her workin g list. After a reviewer accomplishes his/her review process by providing comments and recommendations to the editorin-chief/editor, data fields in Table img_reviewer_process are updated. Especially, attribute ReviewerDecision is assigned a non-null value indicating the submission is removed from the working list of reviewer. Emails are sent out to inform edito r and editor-in-chief that the submission has been reviewed. The last step of the peer review process is the decision-making process conducted by the editorial board. If a subm ission is accepted, attribute Statu s in Table img_lib is changed from to If a submission is accepted with revision, text information of the submission should be revised based on the comments of the editorial board, and then the status of the submission is changed from to If a submission is rejected, the status of the submission is changed from to Once a decision is ma de by the editorial board, the submission is removed from the editor-in-chief/editors In Revi ew list, and an email is sent out to notify submitter the decision on the submission. 3.3 Media Search/View The Media Search/View com ponent provides functionalities to public users to search and view submissions which have been accepted (with revision). This component includes two modules, Media Search and Media Vi ew. Media Search retrieves data that qualify search criteria, and Media View organizes and shows the search results.

PAGE 30

30 3.3.1 Media Search: Extensive Search Options Media Search m odule provides several search options with the aim of improving user experience and search accuracy in DDISMedia. After Media Search receives request from user, it calls the corresponding SQL stor e procedure to retrieve matchi ng records. Extensive search options are listed as follows. Keyword search : This is a generic search based upon the keyword that user input. The keyword search procedure checks attribut es CommonName, Genus, Species, SubSpecies, HostPlant, HostGenus, H ostSpecies, HostSubSpecies, Family, AdditionalDscp, SubmitBy and Photographe rName in the Table img_lib. If a record in database has any at tribute containing the keyword, th is record is included in the result set. ID search : This option allows user retrieve the submission with exact library ID. Advanced search : User may select different terms to customize their search criteria. Search terms include common name, scientific name which is composed of genus, species and sub species, common name and scientific name of host plant, sample type, family, country, state and date range. If user specifies two or more sear ch terms, the search results must meet all search requirements. Browse all : This option allows users to easily re trieve all accepted submissions in the library, and does not require user to set any search conditions. Tag-based search : Tag is defined as the keyword which appears in the common name field or common name of host plant field of one submission. Up to 100 tags with the highest frequency in database are dynamically extracted by the system at each time when user accesses this feature. A tag can be used as a keyword in the Keyword Search option. 3.3.2 Media Browser Media Brow ser is a module which retrieve s and shows complete information of one submission specified usually by a library ID. If a submission has been accepted (with revision), the sole task for Media Browser is to retrieve and exhibit data. If a s ubmission is in the peer review process, extra work is needed before re trieving data: Media Browser checks if a user has access to the submission.

PAGE 31

31 Table 3-1. Data schema of Table img_lib Attribute name Data type Allow null LibID char(17) Unchecked CommonName varchar(125) Checked Genus varchar(75) Checked Species varchar(75) Checked SubSpecies varchar(75) Checked HostPlant varchar(125) Checked HostGenus varchar(75) Checked HostSpecies varchar(75) Checked HostSubSpecies varchar(75) Checked SampleType char(10) Checked Family varchar(75) Checked County char(45) Checked City char(45) Checked State char(45) Checked Country char(45) Checked Longitude char(10) Checked Latitude char(10) Checked CollectionDate Datetime Checked AdditionalDscp Ntext Checked PhotographerName varchar(256) Checked PhotographerInstitute varchar(85) Checked PhotographerEmail varchar(50) Checked Status Smallint Unchecked SubmitDate Datetime Checked SubmitBy char(12) Checked EditorUpdateDate Datetime Checked EditorBy char(12) Checked EditorDecision char(1) Checked EditorComments Ntext Checked CopyrightNote Ntext Checked Table 3-2. Data schema of Table img_media Attribute name Data type Allow null LibID char(17) Unchecked ImgID char(2) Unchecked FileName char(25) Checked FileType char(35) Checked Media image Checked

PAGE 32

32 Table 3-3. Data schema of Table img_editor_process Attribute name Data type Allow null UserID char(12) Unchecked LibID char(17) Unchecked ReceiveDate datetime Checked Decision smallint Checked DecisionDate datetime Checked Status smallint Checked Comment ntext Checked Table 3-4. Data schema of Table img_reviewer_process Attribute name Data type Allow null AutoID Int Unchecked LibID Char(17) Unchecked UserID Char(12) Unchecked CommonName varchar(125) Checked Genus varchar(75) Checked Species varchar(75) Checked SubSpecies varchar(75) Checked HostPlant varchar(75) Checked HostGenus varchar(75) Checked HostSpecies varchar(75) Checked HostSubSpecies varchar(75) Checked SampleType Char(10) Checked Family varchar(75) Checked Comments_Private Ntext Checked Comments_Public Ntext Checked ReviewerDecision Smallint Checked AssignDate Datetime Checked ReviewDate Datetime Checked

PAGE 33

33 Figure 3-1. The ID generation flow chart in Media Submit Figure 3-2. Peer review process in DDISMedia

PAGE 34

34 CHAPTER 4 USER INTERFACE DESIGN The user in terface (UI) is the aggregate of m eans by which people interact with the system. A well designed UI makes the process of using the system effective, efficient and satisfying. In DDISMedia which is a web application, UI especial ly refers to web user interfaces (WUI). It accepts input and provides output by generating we b pages which are transmitted via the Internet and viewed by the user using a web browser program. Some tec hnologies such as JavaScript, Ajax, are utilized to provide real time interactions between users and pages, eliminating the need of refreshing a traditional HTML-based web browse r. This chapter exhibits the user interface of each component in DDISMedia. 4.1 Media Submission UI One design concept of DDISMedia is to prov ide a UI which enables quick subm ission for submitter. Only a small number of input fields in the submission are required. The submission form is composed of six parts: Sample Information form, Location form, Additional Description form, Copyright Notes form, Photographer/Vi deo Producers form, and Upload Media form. The Sample Information form collects sample information. As shown in Figure 4-1, common name and scientific name (including ge nus, species, and sub species) of pest, common name and scientific name of host plant, sample type, and family are input fields in this form. All fields in this form, expect the Sample Type fi eld, are optional because the peer review process is conducted to ensure all information is comp lete after media submission. Sample type is required because editor-in-chief would assign su bmissions to correspondi ng editor according to its values. The value of sample type is in the following pre-defined cate gories: Plant Disease, Insect (Plant), Insect (Non-plant), Plant/Wee d, Management/Physiology/Nutrient, and Livestock. Family filed is usually applicable for submissi ons whose sample type is insect or plant.

PAGE 35

35 The Location form is shown in Figure 4-2. This form aims to collect sample geographic information such as country, state/province, city, county, longitude, and latitude. Country, State/Province, and Media Collection Date are three required fields. Cl icking the calendar icon pops up a small window which helps user choo se and input the me dia collection date. Information in this form is mainly for statistical purposes. Considering the complexity and inconveni ence of inputting longitude and latitude, DDISMedia provides a nice feature to aid user to find their geo-codes. When a user clicks link Get GeoCode in the Location form, a geo-code input panel shown in Figure 4-3 is popped up. This panel is implemented by embedding the Google Ma p API. On the left side of this panel, it is allowed to generate geo-code by clicking Gen erate Geocode button. On ce an address in the Location section is verified by the Google Map API, a pair of longitude and latitude coordinates is automatically fill ed into the correspondi ng fields in the GeoCode section. On the right side of this panel where a Florida map is shown by default, users are able to drag the green star icon on the map to locate any place in the worl d. When the star icon is dragged, the geo-code information in the GeoCode section is updated accordingly. By clicking Use this geocode and close this page button, values in fields Longi tude and Latitude are input into corresponding fields in Figure 4-2. Both the Additional Description form and the Copyright Notes form include optional fields where user input additional descri ption and copyright messages, respectively. The user interfaces for these two forms are shown in Figure 4-4. Photographer/Video Producers form, which is shown in Figure 4-5, is the interface to gather photographer/video producers personal information, such as na me, institute, and email. If the submitter is the person who produces photos or video clips, these fields could be skipped by

PAGE 36

36 checking the checkbox in this form: the system is able to complete related information for the user. Figure 4-6 shows the Upload Media form. By clicking the Browse button, a submitter can select media files from his/her local disk. By clicking link Add and Delete, a submitter is able to add and delete input fields, respectively. A submitter is allowed to decide the number of media in one submission at their will; however, it is strongly recommended to limit the total file size to 50 MB to avoid heavy burden on the ne twork. DDISMedia only accepts image files in jpeg, gif or png format, and vide o files in mpg or wmv format. Clicking the Submit button in Figure 4-6 is the last step in me dia submission. If any required information is missing or an invalid valu e exists in the submission form, a dialogue is popped up to inform submitter that error occurs a nd the submission is invalid. If a submission is processed successfully on the server, a message is returned to the client notifying the success of submission. 4.2 Input Suggestions for Pest/Host Names It is a goal to design a user in terface collecting consiste nt user input, since data inconsistency could jeopardize th e search ability and library qu ality of DDISMedia. However, the length and complexity of th e common name and the scientific name of the pest/host makes it easy for a user to input inconsistent names or invalid input. To solv e this problem, input suggestions are provided to users when input characters are captured. Ajax, a newly developed technology which allows interaction between page s and users without refreshing, is used to accomplish this task. An example of Ajax-based input suggestions is shown in Figure 4-7. As shown in Figure 4-7, immediately after th e characters or ar e captured, a list of options starting with or appears. The user ca n select one of these op tions, and the selected string will appear in the Pest Common Name field. Options are extracted from a pre-defined

PAGE 37

37 list containing over 16,000 names, which is mainta ined by the National Pe st Diagnostic Network (NPDN). Due to the large number of names in this list, this input suggestion feature is enabled only when two or more characters are input. 4.3 Media Review UI The user in terfaces in Media Review compone nt are composed of the following parts: Working List UI, Assignment UI, View Co mments UI, Review UI, and Decision UI. 4.3.1 Working List UI Working list is available to all three user gr oups (i.e. editor-in-chief, editor, and reviewer) involved in the peer review pro cess. It is technically a table containing subm issions waiting for assignment and/or review. Working list for ed itor-in-chief/editor is further divided into Unassigned list and In Review list. Figure 48 shows an example of In Review working list for editor-in-chief/editor. In this table, the first four columns containing basic information such as library ID, submitters name, sample type, and submission date, are identical for all three user groups. The library ID is linked to the me dia viewer (Section 4.4). The Action column, however, is more complex. For editor-in-chief and editors, if it is in the Unassigned working list, only Assign/View Reviewer link is shown in Action column ; if it is in the In Review working list, both Assign/View Reviewer and View Reviewers Comments links are shown in Action column as showed in Figure 4-8. For reviewers, the only link in the Action column is Review. 4.3.2 Assignment UI Assign/View Reviewer link as sh own in Figure 4-8 navigates editor-in-chief/editor to the Assignm ent UI, which helps editor-in-chief to assign editors and al so helps editor-in-chief/editor to assign reviewers for a submi ssion. Figure 4-9 shows an example where the editor-in-chief is assigning editors and/or re viewers for submission -15428. In the ASSIGN EDITOR

PAGE 38

38 section, the editor-in-chief is able to designate any us er within the Assign/Change Editor pulldown list as the editor of this submission. If a submission has not been assigned to any editor, message Media has not been a ssigned. is showed under text ASSIGN EDITOR; otherwise, the assignees user ID and assignment date ar e shown. Currently, a submission can be assigned to only one editor; however, the assigned edito r is allowed to be changed if necessary. Similarly, as shown in Table 4-9, the ASSIGN REVIEWER section is the place to assign a submission to reviewers. All specialists in DDIS are reviewer candidates. Ajax is utilized to provide input suggestions for the Assign a reviewer text field. If a submission is assigned to a reviewer, the assignees name and us er ID and assignment da te will also be s howed on the screen. As shown in the example, submission 08-15428 has been assigned to a reviewer whose name is j Xin and user ID is imglib on Jun. 18, 2008. A submission can be assigned to multiple reviewers. As for the Assignment UI for editors, the onl y difference from the example above is the absence of the ASSIGN EDITOR section since ed itor is not allowed to assign a submission to other editors. 4.3.3 Review UI The Review link in th e review ers working list navigates revi ewers to the Review UI. An example of the Review UI is shown in Figure 410. In this UI, reviewers are allowed to provide recommendations and comments to editor-in-ch ief and editor. There are three options of recommendation: Accept, Accept with revision, and Reject. The Comments to Submitter and Editor text area collects comme nts to both submitter and edit or-in-chief/editor, while the Comments to Editor text area collects co mments that are only viewable by editor-inchief/editor.

PAGE 39

39 If the option Accept with revision is selected by the reviewer, the screen will be different from Figure 4-10 and an example is shown below. As shown in Figure 4-1 1, an extra section is appended to the Comments to Editor text area. By default, submitters original inputs are retrieved from database and filled into the corr esponding fields. Reviewer is allowed to modify common name, scientific name, host common name, host scientific name, and family. 4.3.4 View Comments UI View Reviewers Comm ents link shown in Figure 4-8 navigates editor-in-chief/editor to the View Comments UI, where editor-in-chie f and editors can view recommendations and comments provided by the reviewers. In an exam ple of the View Comments UI as shown in Figure 4-12, a submission with library ID 08-43881 was reviewed by a reviewer on September 30, 2008, whose name is aba zhang and user ID is aba. All information input by reviewer through the Review UI is shown in the COMMENTS message box. Since multiple reviewers can be assigned for one submission, more than one message boxes may exist at the same time. If the reviewers do not provide any comment or recommendation, the text message There is no comment right now. will be shown in the COMMENTS section instead. 4.3.5 Decision UI In Decision UI, editor-in-chief and editors m ake their final decision for a submission. As shown in Figure 4-13, this Decision UI is very similar to reviewers Review UI. A decision can be made from the following three options: Accep t, Accept with revision, and Reject. And again, as discussed earlier, the selec tion of Accept with revision w ould result the UI as shown in Figure 4-11.

PAGE 40

40 4.4 Media Search/View UI The Media Search/View UI is divided into two parts : Media Search UI and Media View UI. 4.4.1 Media Search UI The Media Search UI is shown in Figure 4-14. A s shown in the screen, there are four tabs corresponding to four search options: basic search, media ID sear ch, tag-based search (i.e. tag cloud), and advanced search. The blue tab repres ents the active search option, while gray tabs represent inactive search options. The user interface for Basic Search is s hown in Figure 4-14. The keyword could be any string composed of several characters. The user interface for Media ID S earch is very similar to the screen shown in Figure 4-14: instead of keyword, a valid library ID is needed to get a result for the Media ID Search. A Tag Cloud screen (Figure 4-15) is an exam ple of the user interface of tag-based search option. Tags are arranged al phabetically in the tag box. Tags in bigger font size indicate higher frequencies in the database. It is worth mentioning that a tag is not limited to one word in DDISMedia. For instance, Tomato Spotted Wilt Virus is treated as one tag. Tags are always up-to-date because they are generated ev ery time when the page is refreshed. The Advance Search UI is showed in Figure 4-16. Users can customize search criteria using the following fields: common name and scient ific name of pest/host, sample type, family, country, state/province, and submission date range. If users specify multiple fields, the search results must qualify all conditions simultaneously. 4.4.2 Media View UI The Media View UI organizes and shows search results and are com posed of the following three components: square image viewer, t humbnail image viewer, and media viewer.

PAGE 41

41 The square image viewer is a user interface showing results retrieved by the Browse All search option. As shown in Figure 4-17, each page is designed to show forty images in the order of their submission dates. A page navigation bar is available at the bottom to navigate user to their desired page. Each image has a mouse over to show common name and scientific name of pest. The thumbnail image viewer is a UI to browse search results of the following search options: basic search, media ID search, tag-based search, and advanced search. In the example shown in Figure 4-18, the word tomato is used as the keyword. The left side of the upper tool bar shows the number of search re sults and search conditions, and the right side allows user to set the number of images showed in one page. Be low the tool bar are thumbnail images of search results. The common name and scientific na me of pest is showed below each image. By clicking any image in the square image reviewer or the thumbnail image viewer, detailed information of the corresponding submi ssion is showed in medi a viewer, including text information and all digital media in this s ubmission. Figure 4-19 shows an example of media viewer. If a media is an image, its normal sized copy with a link navigating to the original image is shown in the media viewing panel. If a medi a is a video clip, it is embedded in the media viewing panel and playable. Below the media view ing panel is the area showing all related text information for the submission, such as common na me, scientific name, location, copyright, etc. Two tool boxes are located at the right side of the media viewer. The upper tool box lists the thumbnail images of all digital media in one submission. Clicking any thumbnail image in this tool box would let the correspondin g media shown in the media viewing panel. The lower tool box provides search links implementing simple asso ciated search functions Associated search topics include scientific name of pest, scien tific name of host, photographer, and submitter.

PAGE 42

42 4.5 Interfaces within DDIS The m edia library database of DDISMedia is coupled with DDIS pest diagnosis by embedding a function in DDIS diagnosis user in terface (Figure 4-20). By default, the button Search related topics in DDIS Media Library is disabled; however, it ca n be activated once any character is input in hos t or diagnosis field. For example, T omato is used as the common name of host. By clicking the Search related topi cs in DDIS Media Libr ary button, the Media Search/View component will be called to retrieve and show the media in a pop-up window, whose host common name starts with Tomato.

PAGE 43

43 Figure 4-1. Sample information form Figure 4-2. Location form Figure 4-3. Geo-code input panel

PAGE 44

44 Figure 4-4. Additional description form and copyright notes form Figure 4-5. Photographer/ video producers form Figure 4-6. Upload Media form

PAGE 45

45 Figure 4-7. Example of Ajax -based input suggestions Figure 4-8. Example of In Review working list UI for editor-in-chief/editor Figure 4-9. Example of Assignm ent UI for editor-in-chief

PAGE 46

46 Figure 4-10. Example of Review UI for reviewer Figure 4-11. Example of user interface when se lecting Accept with revision in Review UI

PAGE 47

47 Figure 4-12. Example of View Comments UI for reviewer Figure 4-13. Example of Decisi on UI for editor-in-chief/editor Figure 4-14. Media Search UI

PAGE 48

48 Figure 4-15. Example of Tag Cloud UI Figure 4-16. Advanced Search UI

PAGE 49

49 Figure 4-17. Example of square image viewer Figure 4-18. Example of thumbnail image viewer

PAGE 50

50 Figure 4-19. Media viewer Figure 4-20. DDISMedia embedde d in DDIS pest diagnosis

PAGE 51

51 CHAPTER 5 SUMMARY, CONTRIBUTIONS, AND FUT URE DIRECTIONS 5.1 Summary A study involved the design and im plementatio n of DDISMedia: an online digital media library used in pest diagnosis, research, and e ducation. This web-based system allows users to submit digital media, such as video clips a nd photographic images obtained from the field, clinics or diagnostic laboratories as an impor tant reference for pest diagnosis in DDIS. DDISMedia utilizes a peer review process to ensure high-quality media and authoritative information within this library. The extensive search capability of DDISMedia provides users a good reference to identify and understand pest management. Before designing DDISMedia, potential featur es and requirements for this system were determined, including well-designed graphic user interfaces, reliable media submission process, quality control peer review pro cess, extensive search ability, and integration with DDIS pest diagnosis. The core functional components of DDISMedia contain Media Submit, Media Review, Media Search/View, and Assisted Diagnosis. Media Submit collects media information and implements a stable file uploa ding function. Media Review relies on a peer review process to ensure the qua lity of the digital library. Med ia Search/View provides public users extensive search ability and user-friendly browsers. Assisted Diagnosis embeds DDISMedia into DDIS diagnosis user interface to assist pest diagnosis in DDIS. The following technologies and tools have been used in the implementation of DDISMedia: HTML, JavaScript, Java Servlet, JSP, Apache FileUpload, SQL, JDBC, and Ajax. 5.2 Contributions As specialis ts increasingly depend on multimed ia to help them obtain information and make decisions, DDISMedia currently serves as a useful reference resource for pest diagnosis and identification in DDIS. In addi tion to the textual field information, it is convenient to compare a diagnostic sample visual ly with the archived media in the library.

PAGE 52

52 With the accumulation of media in DDISMedia, it will become an important reference tool for DDIS. The extensive search ability of DDISMedia pr ovides a rich user interface to assist specialists with pest diagnosis. With the exte nsive search ability, th e relationship of pest taxonomy is expressed, which is not availa ble in the DDIS search functionality. 5.3 Future Directions The curren t search capability of DDISMedia has room for improvement. For the current dominant relational database syst em, it is possible to fail to find out the relationship between two associated samples just due to the absence of si milar keyword between them in the database. As one of the solutions to solve the problem, in very small and specified field, people can apply their perception and knowledge in computer preprocessing, such as clustering, or building decision keys in the database to improve the se arch function. However, when the scope grows, the manual work would increase exponentially and might be impractical for implementation. Therefore, future research on improving da tabase technology could be performed. Moreover, implementation of a user interface with the ability to visually show the sample relationship is another interesting topic. For exam ple, a hierarchy or netw ork structure might be one of the relationships among bei ngs in biological chain. By using rich internet technologies, such as Ajax and Flex, it is feasible to implement this feature. In addition, image processing applets can be applied to media viewer. For instance, improving image quality could make images more readable, and emphasizing biological features in an image could assist pest diagnosis and become an improvement in long-distance pest diagnosis. At last, although other diagnosti c labs or organizations might have good-quality media, it is difficult for DDISMedia to connect their data base directly without permission. Therefore,

PAGE 53

53 obtaining data from the search user interfaces of those severs and showing these data in DDISMedia to extend the search scope of DDISMedia is another topic dese rved further research.

PAGE 54

54 LIST OF REFERENCES Bergsten, H ans, 2003, JavaServer Pages, O'Reilly, Cambridge, MA. Crane, Dave, and Pascarello, Eric, 2 005, Ajax in Action, Manning, Greenwich, CT. Flanagan, David, 2006, JavaScript: The Definitive Guide, O'Reilly, Cambridge, MA. Goodman, Danny and Morrison, Michael, and Ei ch, Brendan, 2004, JavaScript Bible, John Wiley and Sons, Hoboken, NJ. Groff, James, and Weinberg, Paul, 2002, SQ L: The Complete Reference, McGraw-Hill Professional, New York, NY. Holmes, Gerald, Brown, Edward, and Ruhl, Gail, 2002, Whats A Picture Worth? The Use of Modern Telecommunications in Diagnosing Pl ant Diseases, Plant Disease, Vol. 84 No. 12, 1256-1265. Hunter, Jason, and Crawford, William, 2001, Java Servlet Programming, O'Reilly, Cambridge, MA. Kurniawan, Budi, 2002, Java for the Web with Servlets, JSP, and EJB, Sams Publishing, Sebastopol, CA. Mahemoff, Michael, 2006, Ajax Desi gn Patterns, O'Reilly, Cambridge, MA. Musciano, Chuck, and Kennedy, Bill, 2007, HTML & XHTML: The Definitive Guide, O'Reilly, Cambridge, MA. Paulson, L.D., 2005, Building Rich Web Applicati ons with Ajax, Comput er, Volume 38, Issue: 10, 14-17. Perry, Bruce W., 2004, Java Servlet and JSP C ookbook: Practical Solutions to Real-world Problems, O'Reilly, Cambridge, MA. Ramakrishnan, Raghu, and Gehrke, Johannes, 20 02, Database Management Systems, McGrawHill Professional, New York, NY. Smith, K., 2006, Simplifying Ajax-style Web development, Computer, Volume: 39, Issue: 5, 98101. Willard, Wendy, 2007, HTML: A Beginner's Guide, McGraw-Hill Professional, New York, NY. Xin, Jiannong, Beck, Howard, Halsey, Larry, Flet cher, James, and Zazueta, Fedro, 1998, Using Digital Cameras and the Internet to Iden tify Plant Insect and Disease Problems, Proceeding of 7th International Conferen ce on Computers in Agriculture. Orlando, Florida.

PAGE 55

55 Xin, Jiannong, Zazueta, Fedro, and Beck, Howard, 2003, A web-ba sed distance diagnostic and identification system for extens ion, EFITA 2003 Conference, 460-465.

PAGE 56

56 BIOGRAPHICAL SKETCH Shiwei Zhang was born in 1981, in Chongqing, China. He received his Bachelor of Engineering degree in com puter science and technology from Nankai University in 2003. He began his graduate study in a concurrent degree program at the University of Florida in spring 2006. After completing the Master of Science degr ee in the Computer Information and Science Engineering Department in December 2007, he con tinued the Master of Engineering degree in the Agricultural and Biological Engineering Department, in the area of information technology under the supervision of Dr. Fedro Zazueta. As a research assistant, his work focused on redesigning and implementation of DDIS, a Di stance Diagnostic and Identification System, which is currently widely used in pest diagnosis by the state of Florida and some international organizations. He also published two academic papers at World Congress on Computers in Agriculture.