UROP Opportunities at W3C

Summer 2000 Opportunities

1. The W3C hosts many meetings for its members and others to further its goal of developing technical specifications for the Web. We have developed a web-based meeting registration system to support this activity. More recently we developed a new web-based membership database. The existing meeting registration system does not currently interface with the member database. We would like someone to write a new meeting registration system which integrates with our member database. Knowledge of PHP3 and SQL is required.
2. The W3C Website is equipped with an access control system which grants access to various parts of the content based on the identity and sponsorship of the user. The user information is kept in an SQL database. We would like someone to write Web applications to help us manage that user database. In particular, some of the access rights would be delegated to official W3C Member representatives at their discretion. Knowledge of PHP3 and SQL is required.

These UROPs will report to Alan Kotok, Manager of W3C Systems & Web Team, with direct support by W3C team members who have intimate knowledge of the structure of the W3C databases and applications.

Queries should be addressed to Alan Kotok.

The following information pertains to past UROPs:

1997-1998

(Revised $Date: 2000/04/14 19:37:56 $ by $Author: kotok $)  -changes to webreq

OLDER W3C PROJECTS and INFORMATION

• Supervisors-Students - Projects from Summer 1996
• UROP Mini-course June 3 to June 5 1996

The World Wide Web consortium is an industry-sponsored organization hosted by three organizations:

• Massachusetts Institute of Technology at the Laboratory for Computer Science [MIT/LCS] in the United States;
• Institut National de Recherche en Informatique et en Automatique [INRIA] in Europe;
• Keio University at the Shonan Fujisawa Campus [Keio/SFC] in Asia.

Our goal is to realize the full potential of the World Wide Web. We are a handful of experts in computing, network protocols, security and related aresas trying to lend order and excitement to the ever-evolving Web. Our work is often technical, but it also involves the relationship of our technologies to society. We are concerned not only with creating new technologies but in guiding their introduction sothey have maximum positive impact. Here at MIT/LCS we are seeking a few very good students to work with our team on some interesting projects. Some of these projects Proposed for this summer are listed here.

GUIs for Expressing Privacy

(Supervisor: Ralph Swick, swick@w3.org, 258-5740)

This is an opportunity to work with architects who are building mechanisms for enhancing personal data privacy for users of the World Wide Web. A piece of this puzzle is the user interface(s) by which a non-expert Web user can configure the system to enforce his or her own wishes. Much experimentation is possible; the UROP candidate will learn a lot about the social implications of graphical user interfaces.

Annotating Web Resources

(Supervisor: Ralph Swick, swick@w3.org, 258-5740)

Standard mechanisms for encoding and transporting descriptive data about Web pages exist and will soon support extended capabilities. Like a library card catalog entry that describes a book in the library collection, these descriptions can be stored and accessed independently of the Web page itself. One use of this facility is to have "write-on card catalogs" where library patrons can leave their own additions, both for their own use and for the benefit of future library patrons. Implementations of such mechanisms and research into their practical impact on shared data spaces is the thrust of this project.

Privacy Watchdog

(Supervisor: Ralph Swick, swick@w3.org, 258-5740)

There currently exist a multiplicity of ways about which users are expected to be aware to configure their systems to protect their personal privacy when browsing the World Wide Web. The number of different Web browsers and other tools leave a non-expert user more exposed than is typically realized. This UROP project will look at developing software assistants, or "wizards" to advise a user how to configure the privacy protection aspects of a system and to watch over a system as it is being used to warn of any potential personal privacy "leaks".

Metadata standards and Software agents

(Supervisor: Ora Lassila, lassila@w3.org, 633-6260)

The World Wide Web has grown into a vast repository of information, and is approaching a stage where automated tools are required to cope with all the data available "out there". We are developing representations -- called "metadata" -- which allow web objects (pages, search services, etc.) to describe themselves in "machine-readable" form. These metadata mechanisms could also be used by "intelligent agents" for exchanging information about about themselves, their capabilities, goals, etc. In this project we would like to experiment with simple software agents which would use emerging metadata standards for communicating as well as for discovering things on the web.

Filtering Sorter

(Supervisor: Philip DesAutels, philipd@w3.org, 258-5714)

Searching on the internet is a black art of boolean operators and secret incantations, and even WITH these, finding the right information quickly is difficult at best. It need not be! As a key member of this world-wide project team, you will work with the cutting edge Web technology developers to design and implement a dynamic hybrid filtering search engine. The ultimate goal of this project is to provide a showcase where the benefits of content labeling for advanced information discovery are clearly demonstrated.

Content Labeling Tools

(Supervisor: Philip DesAutels, philipd@w3.org, 258-5714)

While there are scores of tools for creating Web content, few provide metadata about that content. The goal of this project is to create 'plug-in' tools which will allow content creators to provide cataloging information about their pages. This tool will integrate with popular commercial authoring tools and leading-edge web infrastructure. The UROP candidate on this team will work with the developer of the first Java-based web server and developers from leading commercial web product companies. The results of this project will be made widely available to an eager web community.

UROP Program of W3C during 1996

• Supervisors-Students - Projects from Summer 1996
• UROP Mini-course June 3 to June 5 1996

Older - Projects- (proposed in 1996)

This is a unordered list of some of the projects that can be used by UROPs or everybody else who wants to contribute to the development of the Web. Comments and new ideas are always welcome!

Top-down parser for PICS label lists

Implementation of a simple top-down parser for PICS label lists. Written in either Lex/Yacc, Scheme, Java and/or C, this would have to be very, very simple to understand, and be fully commented.
Proposed by Jim Miller

Signing of PICS labels

Implementation of code to sign PICS labels. Possibly based on existing code to parse a label list, this would take a parse tree, canonicalize the label, sign it, and emit the fully signed version. Might be extended to include third-party signing of digital code.
Proposed by Jim Miller

User interfaces for configuring a PICS browser

Rapid prototyping of several Windows-based user interfaces for configuring a PICS browser with more complex filtering sets than supported by the current reference code or Internet Explorer code.
Proposed by Jim Miller

Constraint model for Web robot

We have a small robot written in C but it does not currently have a constraint model for how to travers the Web. The purpose of this project is to give the robot a "brain" so that it can traverse the Web in an ordered manner.
Proposed by Henrik Frystyk Nielsen

Mail client

Write a email client in C so that you can send and receive mail using your Web browser.
Proposed by Henrik Frystyk Nielsen

todo list

Write a set of resources to manage a todo list. If generic enough this could be used as a replacement for our sysreq stuff, but also for tracking bugs, etc. Todo items come into the list, are assigned to someone (the only one allowed to change its status afterward), and are kept in a history list. Various kinds of display should be provided (per status, per owner, per date, etc).
Proposed by Anselm Baird-Smith

annotations

Write a generic annotation filter, that would decorate documents being served with their annotations. This could probably use the available PICS code.
Proposed by Anselm Baird-Smith

mail access

Provide mail access to existing mail folders: this should allow you to at least incorporate new mails and read them. This might be done through some existing protocol (IMAP, the right way).
Proposed by Anselm Baird-Smith

proxy resource

Depending on how much you rely on the Java runtime, this can take one day to one month. Hopefully the Jigsaw release should come with the most simple version of it (no extra caching than the one provided by the Java libraries), but this can/should be upgraded to more suitable things (ie implement 1.1 caching ?).
Proposed by Anselm Baird-Smith

PICS proxy

Write a client side PICS handling and incorporate it into the proxy resource. This should include a service and a label parser, a way of setting the proxy allowed ratings (at least using the generic resource editor).
Proposed by Anselm Baird-Smith

Calendar

Write a shared calendar resource. This should display the calendar in some nice ways, and allow for its collaborative edition.
Proposed by Anselm Baird-Smith

Automation of Web page maintenance

The W3C team has a large number of content pages. Because the w3c team is distributed internationally it is essential that team members have easy structured access to team project and other content pages. Perl or Python scripts are needed to update and position files for easy access by team members. The successful student applicant, will develop skills in writing such scripts and applying tools and techniques developed by the team or others to the information needs of the W3C team, and (possibly for the larger W3C Membership).
Proposed by Tom J. Greene

Secure, automated remote regression tests

We will soon have a secure authentication infrastructure. Once that is in place we could set up an automated regression testing tool based on a Web server on the test machines. No need to log into each machine to run tests anymore..
Proposed by Phill Hallam-Baker

Interactive Conference Tool

I have this interactive conference tool (in C) which could be developed a lot further, made more reliable etc. It is logically equivalent to IRC but with better controlled dialog structure. Basically this is a turn a weekend hack into a product project.
Proposed by Phill Hallam-Baker

Core PEP library design

Possibly working as a plug-in module for Internet Explorer to integrate PEP proxy code directly into a client. Learn about COM and designing extensible architectures. May extend to user-interface considerations. Or, take the current PEP proxy and build interesting applications: a cash register, a shopping cart protocol, a session-tunneling layer, demographics, etc.
Proposed by Rohit Khare

Simulating a micropayments scheme

Both the design of a simple "reputation-based" currency for valuing web pages, and a PEP encoding into the JEPI framework. Even without cryptographic "cash" aspects, it is a powerful combination of voting and demographics
Proposed by Rohit Khare

Anyone at all interested in Web Security

I'll be glad to work around their interests: in channel-layer stuff (SSL, PCT), dig signatures, practical investigations (safety of downloadable code languages), etc.
Proposed by Rohit Khare

Advanced Projects

These are things that you can start on when you have gotten your fingers wet in the Web world.

• Constructing an operational semantics for the Web protocols would be v. useful.
• There are a lot of interesting things to do with caches and proxies. A distributed cache protocol would be very interesting.
• In the Industrial Strength Web area a clustering protocol (ala VAXCluster) would be a step to solving many reliability problems.
• Distributed and Persistent Cache