CSCI 577a Projects – Fall 1999

  1. Doheny Library Virtual Tour
  2. Viewing Utility for Oversized Scanned Paper Objects
  3. Aggregated List of Full-Text Journal Resources
  4. Digitizing & Access to Chinese, Japanese, or Korean Materials
  5. Daily Summary of the Japanese Press
  6. URL Link Checker/Confirmation Utility
  7. Hispanic Digital Archive Enhancements
  8. Managing Multimedia Databases for Instruction and Research
  9. ILL/OCLC Record Retrieval to Manage Circulation Process
  10. Serials Invoice Loader
  11. Automated New Books List Generator
  12. Chicano/Latino Serials Microfilm Collection on the Web
  13. WEBCAT Extension to Third-party Databases
  14. Vacation/Sick Leave Tracking System
  15. Business/Reference Q & As
  16. Bookstores On Line Virtual Calendar
  17. Bookstores On Line Email & Student-Student Communications
  18. Bookstores On Line Site and Ad Management
  19. MBASE–GroupSystems Project
  20. Electronic Process Guide Authoring tool
  21. Use-Cases to Test Cases
  22. SE Tools InfoBase
  23. MBASE Deliverables Manager
  24. MBASE Guidelines Manager
  25. MBASE Projects Archive
  26. Dashboard Agents and Extensions

 

Project # 1. Doheny Library Virtual Tour
Client Name Matt Gainer
Client Contact gainer@usc.edu; 740-8832
Project Sub-domain Special Collection Access
Project Description Doheny Memorial Library (DML) will soon undergo a seismic retrofit project that necessitates closure of the library from the end of Fall semester 1999 until early Spring semester 2001. The Digital Conversion Production Center (DCPC) has been involved in an ongoing project to visually document the entire library prior to the start of construction. The DCPC has created an archive of hi-resolution digital photographs and low resolution QuickTime VR panoramas that show the library as it exists today.

This project will use the DCPC’s images of DML, along with selected historical images and documents to create a virtual tour of the library for USC’s web pages. Some features might include:

- Linked panoramic QTVR environments.

- spatial indexing of photographs, QTVRs, and documents to floor plans

- A template for organizing and displaying Web-Cam and/or Time-lapse imagery of areas that are largely impacted by the retrofit.

- A template for organizing and displaying before and after images of various locations throughout the library.

   
Project # 2. Viewing Utility for Oversized Scanned Paper Objects
Client Name Matt Gainer
Client Contact gainer@usc.edu; 740-8832
Project Sub-domain Multimedia Archive
Project Description Through a partnership with the Huntington Library, USC has scanned the Huntington’s complete sets of El Clamor Publico and the Los Angeles Star, the first Spanish and English language newspapers in LA. The file sizes for the scans range from 30MB to 85MB, and in total they occupy approximately a terabyte of disk-space.

The first part of the project will involve outlining an ideal viewer for the scans. Next, students will research existing methods of achieving lossless compression and viewability of oversized scanned objects via the internet. If an appropriate solution for viewing does not already exist commercially, students will define the characteristics and components necessary to build a viewer.

   
Project # 3. Aggregated List of Full-Text Journal Resources
Client Name Lynn Sipe
Client Contact lsipe@usc.edu; 740-2929
Project Sub-domain Reference Services and FAQ
Project Description Through ISD's various electronic databases we provide access to several thousand full-text journal resources, as packaged by various vendors, such as in UMI's ProQuest Direct or in IAC's Expanded Academic Index. While these, and the other vendors who products we acquire, provide us with a list of the journal titles covered in each of their products what we lack is an omnibus list that is searchable from the Electronic Resources page.

The creation of an omnibus list of aggregated full-text journal resources is more complicated than it might at first sound, as it is more than a question of just merging the individual lists. We need a methodology to keep the omnibus list current (resources are deleted and added on an on-going basis). Obviously, the omnibus list needs to indicate which journal is covered in which resource aggregation, which will frequently be more than one.

   
Project # 4. Digitizing & Access to Chinese, Japanese, or Korean Materials
Client Name Ken Klein
Client Contact kklein@usc.edu; 740-1772
Project Sub-domain Multimedia Archive
Project Description The East Asian Library periodically receives sizeable donations of Chinese, Japanese, or Korean books and is then faced with the prospect of having to catalog the gift volumes as an add-on responsibility to its regular acquisition schedule. The development of scanning technology now makes it practical to use portable scanners for making an initial inventory of large collections. We would like to see if it is possible to streamline the use of such scanned images (of title pages or colophons) for processing large gift collections, either in preparing a list of titles or, better, for searching the OCLC or RLIN database for cataloging copy. As each language will pose it’s own challenges, team may select the language of the materials they wish to focus on. This project will require the identification and use of OCR software in the selected language and the mapping elements of a scanned title page or colophon into a library-prescribed set of fields.
   
Project # 5. Daily Summary of the Japanese Press
Client Name Janice Hanks
Client Contact jhanks@usc.edu; 740-1773
Project Sub-domain Selective Dissemination of Information
Project Description Background: The American Embassy in Tokyo for years has summarized in English selected articles appearing in the Japanese press and daily published these summaries. Beginning in March of this year they stopped the paper publication and began to send an electronic copy via email attachment. In order to make these available to the USC community I started to set up a master web page and convert them to web documents. It quickly became apparent that I could not easily keep up with the volume of material I was receiving.

Description: Design a system, which would daily [except Saturday and Sunday], convert the Daily Summary of the Japanese Press into an HTML document. The program would then link it appropriately to a master web page. The linking on the master page would be by date. In other words a user could chose a summary for particular day to view. It should also be possible to search text of all the daily summaries by keyword and have a list of hits generated. Within the document the keyword would be highlighted.

See sample press release

   
Project # 6. URL Link Checker/Confirmation Utility
Client Name Nava Herman
Client Contact navaz@usc.edu; 740-0785
Project Sub-domain COTS Package Extension
Project Description As everyone knows, broken links are a major frustration to everyone using the web and a time consuming workload for those supporting access to web resources. We need a utility to inspect and ensure the continued availability of our outside databases, using some kind of configuration file listing the URL's and an expected string (or set of strings) at that URL. The script will do the following:

1. Given a configuration file of URL's and string pairs, the script will check the existence of the given URL and the presence of the specified strings within them.

2. In case of success the script will do nothing.

3. If it fails:

A. Because the URL no longer exists - it should sent an e-mail to a designated person notifying them of that.

B. Because the URL exists, but the search strings specified are not present - again, it should sent an email notifying a designated person/group, but this time the message will contain information about what *does* exist in that URL.

   
Project # 7. Hispanic Digital Archive Enhancements
Client Name Barbara Robinson
Client Contact brobinso@usc.edu; 740-8609
Project Sub-domain Multimedia Archive
Project Description This project draws on the accomplishments of last year in the ongoing development of a web accessible Hispanic Digital Archive (HDA), for organizing, searching and displaying a collection of primary source materials related to the study of Hispanics in the U.S., Latin America and Spain. The principle goals this year are to enhance and extend certain key features and capabilities and explore new issues related to four main areas, which will allow the librarian/client to begin creating a viable database for the Sirsi WebCat Gateway. These four main areas are: 1) functionality (e.g. for metadata ingest and end user access); 2) security (e.g. for protection of images and for certain key fields on the metadata form); 3) editing (e.g. for modification of data in pulldown lists and other metadata fields); and 4) quality control (e.g. for obtaining accurate data and provision of feedback on errors to database inputters). Additional specificity about these will be provided during win-win negotiations between the librarian/client and the team. A new and unique enhancement requested this year, which relates to several of the four areas of focus described above, is the adaptation to the HDA of an existing multilingual electronic thesaurus * ( subject headings in English, with translations in Spanish and Portuguese, available to us either via FTP or on the Web), to be used for verification and automatic creation of standard metadata subject descriptions, eliminating the need for high cost, labor intensive manual inputting, and minimizing errors, as well as allowing consistent access points for data retrieval via the end user/public interface, using any one of the three languages of this thesaurus. Another key enhancement is the development of the media manager, especially for associating and disassociating images with the metadata records and for electronic watermarking of images. The team is encouraged to explore the use of JAVA Swing for the HDA metadata ingest form and to investigate issues and features of BRS and Sirsi for the end user/public interface. * (Permission has been granted to this librarian/client by the copyright owner of the electronic thesaurus for exclusive use in this CS577 project)
   
Project # 8. Managing Multimedia Databases for Instruction and Research
Client Names Karen Howell & Robert Doiel
Client Contacts khowell@usc.edu, 740-2933

rdoiel@usc.edu, 740-2925

Project Sub-domain Multimedia Archive
Project Description Like many other research universities, USC is using new technologies and products to provide a digital environment for teaching, learning, and research. When faculty members create course websites, they often want to provide access to digitized multimedia files for students and researchers to examine. This project will explore the feasibility of using ISD's SIRSI gateway to the BRS database system to support faculty members creating multimedia databases for instruction and research.

Solving this problem would support the curriculum and the scholarly research process by enabling faculty and students to digitize and use the primary resources in USC's collections. This project is an exploration of making audio and video collections at USC accessible, similar to the way IDA/ISLA makes images and geographical information accessible to researchers.

   
Project # 9. ILL/OCLC Record Retrieval to Manage Circulation Process
Client Name Joyce Toscan
Client Contact jtoscan@usc.edu; 740-2931
Project Sub-domain Distributed Borrowing
Project Description The Global Express Department handles the University’s interlibrary loan requests for copies and books from other libraries around the world. It requests approximately 12,000 items each year and received an additional 15,000 requests from other libraries.

The Department is highly automated and accesses major international bibliographic utilities to locate, order and keep records of its transactions. One of the major utilities is OCLC. The scope of the project is to develop an automated method to retrieve specific data from the OCLC files, transfer these data to the University Library’s SIRSI (integrated library system) database, and use its protocols to manage the checkout, overdue, recall, and billing operations for all interlibrary loan returnable borrowing and lending transactions.

   
Project # 10. Serials Invoice Loader
Client Names Susan Scheiberg & Chris Sterback
Client Contacts scheiber@usc.edu; 740-7355

sterback@usc.edu; 740-6913

Project Sub-domain Data Migration
Project Description Among its many duties, the Serials Acquisitions Team is responsible for purchasing well over $3,000,000 worth of serial publications for the University campus. The majority of titles are purchased and renewed through domestic and international third-party subscription agents. To track funds and create an audit trail, the payments for each title received through a subscription agent must be paid for through USC’s Administrative Information System (AIS) accounting system and Sirsi, the library’s integrated library system. Due to the size and complexity of the invoices, payment in the Sirsi system has been sporadic at best.

To ease the manual burden of recording payments and price adjustments in Sirsi, we would like to propose the development of an invoice loader which can manipulate data from the subscription agents and Sirsi records to successfully automate the invoicing process. This will greatly enhance our ability to track and audit serials payments, and free up the Serials Acquisitions Team to focus on other, more customer-oriented tasks.

   
Project # 11. Automated New Books List Generator
Client Name Deb Holmes-Wong
Client Contact dhwong@usc.edu; 740-2867
Project Sub-domain Selective Dissemination of Information
Project Description An individual will be able to sign up on a web form and request that a list of newly acquired books be sent to her on a regular basis. This list is generated from ISD's SIRSI database and should contain enough information from the database for her to determine if she wants to go get the book. The list should come via email in a format that is easy for her to read. She should be able to sign up for multiple subject profiles if she desires. To see a functional example of this type of service visit http://www.marshall.usc.edu/library/forms/subnewbooks.html , this is limited to one subject.

The site should require very little administrative work.

   
Project # 12. Chicano/Latino Serials Microfilm Collection on the Web
Client Name Barbara Robinson
Client Contact brobinso@usc.edu; 740-8609
Project Sub-domain Multimedia Archive
Project Description The ISD's Boeckmann Center for Iberian and Latin American Studies has a voluminous and growing collection on microfilm entitled Chicano Studies Library Serials Collection of periodicals and journals published in the U.S., in Spanish and English, from 1855 to the present. It represents an important source of information and insight into the Hispanic communities nationwide, with extensive coverage of the West Coast. Every year a new section is acquired with about 40-50 additional reels of microfilm containing many new titles. (At present there are 15 sections, with a current total of 512 reels with about 400-500 individual titles). Cumbersome printed lists provided by the publisher of the contents of each section are the primary access to the titles. Although, some of the section contents are also displayed on a Boeckmann Center created web site at URL http://www.usc.edu/isd/locations/international/boeckmann/serials.html . This somewhat primitive solution only reproduces the printed lists and does not allow a flexible and efficient means for identifying the reel contents by title and the geographic and date relationships, nor allow cost-effective updating of the data as new sections arrive. The goals of this project are 1) to develop a searchable web-based database to allow students and researchers to find, browse and display by various attributes (title, city, state, year(s), language, section/reel nos., location); 2) secondly, to create a user-friendly web ingest system (for use on both Mac and PC) to allow efficient input of new reel content data and viewing and editing of existing data for quality control, by staff or student assistants with minimal or no web editing skills; and 3) finally, to explore and identify other potential future enhancements which might include, but are not limited to, fulltext access for the titles in public domain, and the addition to database of related Latino serial titles on microfilm acquired or filmed by USC that are not part of this Chicano collection.
   
Project # 13. WEBCAT-Extension to third-party databases
Client Names Rick Lacy/Dennis Smith
Client Contacts lacy@usc.edu, 740-8823
smith@rcf.usc.edu, 740-2863
Project Sub-domain COTS Package Extension
Project Description Within the framework of Sirsi's WEBCAT and the university databases maintained within it, there is a cgi-based approach for generating unique identifiers for objects within a given database and assigning to any object a "process" URL that uniquely refers to that object. These "process" URL's themselves reference cgi calls upon a database. They are the basis for retrieving not only the object, but also material associated with the object (e.g. metadata). This approach can be used in a number of productive ways -- e.g., generating web pages on the fly which include both the objects and various associated information. Among many other things, it clearly could form the basis for an alternative and much more powerful and efficient electronic reserves system with the potential to generate reserve lists, class syllabi, etc on the fly. This general approach, although currently adapted to WEBCAT and the databases there, is in principle extensible to other environments and databases -- perhaps most especially third party databases that are not part of WEBCAT, such as those the university already licenses and makes available. For example, just following out the simple "electronic reserve" concept, one would be able to build a create a much larger base for an electronic reserve system and in one stroke extract substantially more value for the user out of the resources we are already licensing.

The project therefore is to explore a significant sample of third party databases that we license and determine what the issues and problems would be in extending the current approach to these databases. There are many areas of concern, including:

  • are there structural properties of the systems architecture or software that are problematic
  • is there a basis/means for constructing a unique identifier for each system
  • are there general problems with the cgi approach for that system

Also, having reviewed a number of other databases, and found this approach workable, the greater question is how general the solutions can be. For example, do we have to write substantially different cgi scripts for each database or can we write -- if not common scripts -- structurally similar ones.

   
Project # 14. Vacation/Sick Leave Tracking System
Client Name Sharon Haymond
Client Contact haymond@usc.edu; 740-3034
Project Sub-domain Activity Monitoring and Control
Project Description ISD’s Human Resources needs to track the sick leave and vacation records for over 350 employees; data should be efficiently and accurately gathered from individuals, verified by supervisors, and maintained in confidential files by Human Resources. Individuals should have up-to-date access to information on the amount of vacation/sick leave they have accrued; HR should be able to alert individuals and supervisors when too much vacation has accrued or sick leave/vacation days are overdrawn. The system must be capable of handling differing profiles/policies for each category of employee, based on years of service.

For details on the project see this document.

   
Project # 15. Business/Reference Q & As
Client Name Julie Kwan
Client Contact jkwan@usc.edu; 740-4420
Project Sub-domain Reference Services and FAQ
Project Description Libraries have traditionally provided professional librarians and other trained staff to offer one on one assistance to users who are trying to find information or to solve information problems. Within a given subject discipline, there are typical information problems and solution strategies which are asked over and over again. We would like an automated way to extend reference services to allow Marshall School of Business students to be more independent and proactive in solving their own information queries. Other attempts to provide automated reference service have been done for previous CS577a classes during the Fall Semester 1997, including General Library FAQ’s, the Virtual Education Reference Assistant, and the Virtual Business Reference Assistant. Another approach has been taken by Ask Jeeves (http://askjeeves.com). A number of companies are using question-matching-answer strategies to provide customer support. One possible way to implement the system would be as follows: The student would submit a question to the system. The keywords in the question could then be matched to records in a database, and the system could return a list of documents with suggested solutions. These documents could consist of the question with 1 to 4 strategies for finding an answer. The student would then select the question that best matches his/her needs. The system should require minimal staff time. It should also provide a mechanism to periodically review the content of answers and to handle questions that are new to the system.
   
University Bookstore

USC and every other independent bookstore would like to posture itself to avoid takeovers by the big, online book companies. A consortium of independent bookstores, with USC leading the way, would like create a web-based portal that gives university bookstores an edge on the internet by providing useful services to students at each of these colleges. The site must give the bookstores the ability to compete in the web business: Students must want to use the services, thus providing a highly focused market for each college to advertise to their own students only. This would be accomplished through the customizable sites for each college that wants to participate. Initially, a consortium server would provide all server facilities. The features that are desired of the customizable web sites are 1) A Virtual Calendar; 2) Email and Student to Student Communications; and 3) Site and Advertising management.

While this will eventually be an integrated project, three different 577a projects have been identified and are being offered separately

   
Project # 16. Bookstores On Line Virtual Calendar
Client Name Steve Mosher
Client Contact mosher@usc.edu; 740-1059
Project Sub-domain Interactive Communication
Project Description This is an organizer/calendar interface that automatically populates the user’s calendar with academic and athletic events specific to their school after they register. They would have the ability to add events, classes, meetings, holidays, etc; to their calendar online as well. All their data would be hosted on the consortium remote server so they could access their calendar through the use of a login and password from any computer that has internet connectivity. There should also be the ability to have a local (student computer) based version so the student does not have to be online all the time. This will require synchronization capability. All versions of the calendar MUST have banner-ad space allocated and able to be kept current. The Consortium is willing to buy instead of make, but there is still a great deal of specification and design, development and integration work.
   
Project # 17. Bookstores On Line Email & Student-Student Communications
Client Name Steve Mosher
Client Contact mosher@usc.edu; 740-1059
Project Sub-domain Interactive Communication
Project Description Provide free email service available through the web similar to Hotmail AND remote/offline processing similar to Eudora. It would need to have a good interface and easy to use. A remote address book with names numbers and email addresses is also desired. Other student-student communications mechanisms are desired, including chatting abilities, instant messaging and news-groups with registered friends from this site. All parts and features of this project must be so easy to set-up (and use) that even those with little or no computer literacy. All of the features MUST have banner-ad space allocated and able to be kept current. The Consortium is willing to buy instead of make, but there remains a great deal of specification and design, development and integration work.
   
Project # 18. Bookstores On Line Site and Ad Management
Client Name Steve Mosher
Client Contact mosher@usc.edu; 740-1059
Project Sub-domain Activity Monitoring and Control
Project Description A backend system that allows an administrator from each university to go into their customized backend where they can implement an ad management system for their school. The Site Management portion must handle a registration process where the student identifies the school he/she attends. The Ad Management backend must then recognize that person when they connect and make sure they receive ad banners at the top of their pages that relate only to their university’s bookstore. Clicking on these pages would lead back to their bookstore homepages on their local campus web servers. The Ad and general Site Management systems must function generally on its own after created.
   
Project # 19. MBASE–GroupSystems Project
Client Name Paul Grünbacher
Client Contact gruenbac@sunset.usc.edu;
Project Sub-domain COTS Package Extension
Project Description The MBASE (Model-Based Architecting and Software Engineering) approach aims at identifying, avoiding and overcoming model clashes among product models, process models, property models, and success models in software projects.

Model clashes are often a result of lacking communication among stakeholders. Collaboration Technology facilitating the negotiation process provides a means to avoid and overcome model clashes. Group Decision Support Systems (GDSS) provide a set of software tools supporting a wide range of group processes (e.g., idea generation, issue organization, evaluation and analysis, decisions and policy planning, alternative analysis, and surveys). A popular example is Ventana GroupSystems (http://www.ventana.com). The product allows participants to interact through a network of personal computers either face-to-face or dislocated. Important attributes of a GDSS include structured interaction, parallel contribution, anonymity, group memory, and an intuitive user interface.

The goal of this project is to explore possibilities to integrate GroupSystems into the MBASE framework:

GroupSystems candidate processes in the Inception phase:

  •  
  • Risk Driven Analysis (using Simplifiers & Complicators technique)
  •  
  • Domain Analysis: Support negotiation of system capabilities
  •  
  • Success Analysis: Support the requirements elicitation process following the WinWin negotiation model
  •  
  • Product Analysis: Assess & Analyze COTS Products
  •  

GroupSystems candidate processes in the Elaboration phase:

  •  
  • Risk Driven Engineering
  •  
  • Product Engineering (COTS selection)
  •  

Traceability

We need to develop strategies allowing traceability between negotiation results and other artifacts in the software development process (e.g., negotiated capabilities and the system architecture; WinWin agreements and use cases). Bridges to the USC WinWin negotiation groupware and OOA/OOD CASE tools (e.g., Rational Rose) shall be implemented.

EPG Update

Development of EPG documents describing best practices for group negotiation processes

   
Project # 20. Electronic Process Guide Authoring tool
Client Names Marc Kellner & Nikunj Mehta
Client Contacts mik@sei.cmu.edu

mehta@sunset.usc.edu; 740 6504

Project Sub-domain COTS Package Extension
Project Description The Software Engineering Institute has developed technology for producing Electronic Process Guides (EPG) for software processes. As a collaborative project, USC-CSE has recently developed an EPG for the MBASE approach followed by the students in the CS 577a course. This project involves designing and developing a prototype authoring tool for the EPG as early as the end of December 99.

The EPG technology is in experimental use and is in the form of various Java applications, one of which obtains information from a process database. These tools produce a set of HTML pages containing descriptions of all the process elements. The primary objectives of the tool would be to deliver an easy authoring interface to define and maintain the process information and to provide integration of the various EPG tools.

The tool designed by the student teams will be taken to completion at SEI after the end of the semester. Marc Kellner would serve as a remote client with Nikunj Mehta acting as surrogate client for the project.

   
Project # 21. Use-Cases to Test Cases
Client Name George Huling
Client Contact g.huling@ieee.org; 805- 381 4936
Project Sub-domain CASE Tool
Project Description A lot of interest has been shown in generating test cases that are a random sample of a measured or assumed usage profile. For example, telephone companies do this routinely. If the profile is reasonably close to reality, the test results can be used to estimate user experience (satisfaction) statistically. In 577a, user interactions with the system will be described by Use Cases (narrative descriptions, structured text descriptions and Use-Case Models).

This project is to produce a system that will take detailed use case documentation, add usage profile (frequency) information, and generate test cases that are a random sample of the use case plus profile data.

   
Project # 22. SE Tools InfoBase
Client Name A. Winsor Brown & CSE staff
Client Contact AWBrown@sunset.usc.edu; 740-6599
Project Sub-domain CASE Tool
Project Description The sources of information about tools are varied and sometimes of questionable validity; tool versions change frequently.  The idea is to come up with a web-based (and therefore can be hosted over an intranet) system, which can display information on the tools either in use or for which someone in the organization has first hand knowledge.  The information needs to include automatically updated, traditional and non-traditional sources.  Traditional sources include User Manuals, Reference Manuals, and supplier’s FAQ type information that either came with the product or is available on the WWW.  Non-traditional sources include user group information like FAQs, news group threads, personal experience, and trade magazine reviews and articles.   Information from non-traditional sources should be reviewed, perhaps by local technology experts or gurus, before it is made available to everyone along with a rating and/or comments from the reviewer.

Most software development organizations use many tools and have turnover.  A “refereed” information base should help new people to an organization or a project come up to speed with the tools quickly.  Alternatively, this information should help when a new-to-the-project tool is needed.

This project should result in an operational system, hosted on a CSE server.The initial information in   the system should be about the tools used in 577a.Certain components might be suitable for feasibility study as add-ons, but the core capability should be provided and be made operational.

   
Project # 23. MBASE Deliverables Manager
Client Name A. Winsor Brown & CSE staff
Client Contact AWBrown@sunset.usc.edu; 740-6599
Project Sub-domain Multimedia Archive
Project Description The MBASE guidelines call for a large (both in depth and breadth) array of documentation. To be useful and portable, this documentation needs to be formatted and installed according to strict standards (to be supplied). These storage issues are further complicated by concerns over file formats and the desirability to distribute the documents over the web. These concerns create a significant overhead on the project groups. The goal is to have a highly consistent, web-centric (or equivalent hyperlink-like) documents from each group.

Toward the twin ends of easing MBASE bookkeeping for the project groups, and further standardizing their creations, this project will create a tool for managing MBASE deliverables. There are several possible approaches to this, but minimum features include version control, automated organization of files and a friendly interface. The final implementation might well range from a text-based application to graphical or web-based approaches.

One vision of the project's output is an application that presents the MBASE outline and provides a number of functions on it such as:

• review a section (perhaps with a commenting feature)
• modify a section (with revision control)
• create a new section
• delete a section
• archive all sections
• import or export sections (to various formats such as PDF, HTML)
• print sections
   
Project # 24. MBASE Guidelines Manager
Client Name A. Winsor Brown & CSE staff
Client Contact AWBrown@sunset.usc.edu; 740-6599
Project Sub-domain Process Management Tools
Project Description As CSE learns more about MBASE and best practices in both industrial and student applications of MBASE, changes have to be made to the guidelines.  Changes in some variants of MBASE do not impact other variants of MBASE, but other changes, and all fundamental changes, will impact other variants of MBASE. This project is to automatically generate tailored/customized MBASE guideline documents (with special [high-light] formatting for classes of information and internal links as appropriate), while preserving the evolution history the MBASE guidelines.
   
Project # 25. MBASE Projects Archive
Client Name A. Winsor Brown & CSE staff
Client Contact AWBrown@sunset.usc.edu; 740-6599
Project Sub-domain Multimedia Archive
Project Description Develop and populate from previous year's projects a multi purpose data/information base for the CSCI 577 MBASE projects.The projects should be well organized and cross-connected where appropriate; a user friendly interface is absolutely required, where users might be researchers extracting information, new 577 students wishing to browse project artifacts, or CSE affiliates wishing to learn more about MBASE via exploration of artifacts and MBASE documentation. Each of the projects themselves and their documentation should be more user accessible and cross-connected.Some of the purposes that the Archive is to be put to include data-mining for researchers in process or cost estimation areas, a corpus of experience for analysis leading to identification of best practices, and learning about MBASE.
   
Project # 26. Dashboard Agents and Extensions
Client Name A. Winsor Brown & CSE staff
Client Contact AWBrown@sunset.usc.edu; 740-6599
Project Sub-domain COTS Extension
Project Description Dashboard is a tool present project management information based on analysis of artifacts being produced for a project, like an MBASE 577 Project. Most configurations of Dashboard are hand crafted because of the large variations in projects, but MBASE provides a unifying framework for projects and thus set up of Dashboard could be assisted by tools. One obvious extension is a Project Parameter (INI.file) editor. Dashboard derives project status information through agents that look at the artifacts of the projects. Agents for MBASE would gather information like the number of 1) pages of MBASE document packages, 2) system responsibilities, 3) requirements, 4) system objects, 5) tests, 6) inspections, 7) defects, 8) reviews, 9) changes, and 10) source lines of code.