HP Internet Philanthropic Initiative – Next Generation Internet

Interactive Discussion Forums

Area: Enable new generation of applications

April 1998

Proposal submitted to the Hewlett-Packard European Grants Programs in the
HP philantropy program


Principal investigator

Giuseppe Attardi
Dipartimento di Informatica
corso Italia 40
I-56125 Pisa
tel: +39 (50) 887-244
fax: +39 (50) 887-226
mail: attardi@di.unipi.it
Giuseppe Attardi is professor of Computer Science at the Dipartimento di Informatica, where he currently teaches Computer Graphics.
Prof. Attardi is involved with Internet both at the local level, as responsible of the center SerRA of the University of Pisa, and at the national level, as a consultant in the design of the italian national research network GARR.
Prof. Attardi is responsible for the national Web Cache service.
He is member of a working group of the Italian Ministry of Communications on the initiative for Internet and the development of the information society.
He has been visiting scientist for three years at the MIT Artificial Intelligence Laboratory, where he developed Omega, a calculus of descriptions for knowledge representation based on taxonomies of concepts and participated to the development of the first graphics window system of MIT.
He has been senior visitor at the International Computer Science Institute in Berkeley and at the Sony Research Laboratory in Paris.
He has been project leader of the ESPRIT project P440 (MADS), project CHAMELEON, and group leader in projects APHRODITE, ITHACA, TROPICS, PoSSo and FRISCO.
He has worked on actor languages and concurrency, and developed ECoLisp, an Embeddable Common Lisp.
He is active in the development and implementation of object-oriented languages, including CLOS as part of ECoLisp.
In the ESPRIT project PoSSo, he has been responsible for the development of CMM (Customisable Memory Manager) a dynamic memory management system for C++.
Prof. Attardi is an editor of Computational Intelligence and has served as member of several program committes, including IJCAI, ECAI, ECOOP and KR.
Prof. Attardi is member of the board of directors of the Java Italian Association (http://www.jia.it).
Director of the Dipartimento di Informatica:
Prof. Franco Turini (turini@di.unipi.it)
Maria Simi (simi@di.unipi.it)
Maria Simi is professor of Computer Science at the Dipartimento di Informatica. She participated to the design and development of Omega and she is an expert in taxonomic reasoning and data retrieval. She leads the development of CompAss
Stefano Suin (stefano@unipi.it)
Stefano Suin is technical director of SerRA. He is national coordinator for the national group on USENET News.
Blasco Bonito (bonito@unipi.it)
Blasco Bonito is member of the SerRA team responsible for multimedia protocols and services. Bonito is a member of RIPE.
Sergio Di Marco (dimarco@di.unipi.it)
Sergio Di Marco is a member of the Agents Group at the Dipartimento di Informatica. He developed a Java framework for developing GUI based on a constraint language. He is involved in automatic classification of Web documents.
Davide Salvi (salvi@di.unipi.it)
Davide Salvi is a member of the Agents Group at the Dipartimento di Informatica. He is involved in automatic classification of Web documents, and in particular ranking of categories.
Antonio Gullì (gulli@di.unipi.it)
Antonio Gullì is a member of the Arianna project, developer of Discovery, a suite of tools for guiding the Web search. He is involved in the activity on categorization and summarization of Web documents.
Domenico Dato (dato@di.unipi.it)
Domenico Dato is a member of the Arianna project, developer of Discovery, a suite of tools for guiding the Web search. He is involved in the activity on categorization and summarization of Web documents. He is a radio amateur and a Windows CE programmer.
Antonio Cisternino (cisterni@di.unipi.it)
Antonio Cisternino is a system programmer and Java specialist. He developed the CompAss system for composing plans of study interactively on the Web. He is developing a 3D graphical monitor in Java for the RoboCup.


The World Wide Web has evolved into an impressive open structure for sharing information. The Web is driven by the simple metaphor of navigation. This metaphor however leaves Web users mostly alone with occasional help from tools with limited intelligence. It is essential to develop the Web towards supporting interpersonal communication, allowing interactions with other subjects rather than only with objects. This entails developing simple interfaces for communication services and extending the means to access the Web through new devices, in particular hand held and wireless devices.

In a cooperative effort with Italia On Line, and with the support of equipment provided by the HP INTERNET PHILANTHROPIC INITIATIVE Next Generation Internet, our team has participated in setting up the Arianna search engine, which handles the largest collection of documents in the italian Web. We are currently working on improving such tool by adding a facility of automatic classification of such documents in order to provide means to retrieve documents, which involve selection by subject. See paper Categorization by Context by G. Attardi, S. Di Marco and D. Salvi.

Besides improving the means to navigate and explore the material on the Web, we wish to experiment new means to allow people to use the network to establish communications with other individuals. This entails overcoming the limitations of current on-line chat systems which provide so vast spaces that a user gets immediately "lost", feeling as a single entity within a dishomogeneous, anonymous, unsorted, casually arranged crowd of persons trying to establish human communications.

Our idea is to build systems that enable people to engage in fruitful, purposeful and structured discussion on well-specified themes.

The aims of the project are: to build services for communication and discussion; to develop tools to facilitate interaction among users; and to extend network access to a new class of mobile users.

The first two goals will be achieved through a virtual space on the net where people can join thematic communities and participate in either live or deferred discussions using rich communication media such as audio and video. The third goal will be achieved by building wireless local area networking (LAN) connectivity for handheld devices.


The objective of this project is to experiment with an infrastructure for a Networked Community, focusing in particular on tools for Interactive Discussion Forums. Interactive Discussion Forums are a mean to involve a community in an important issue for the community and reach decisions about how to address the issue. A forum is a form of structured discussion, more purposeful than informal communications, such as chats or mailing list, and has an important role in achieving a sense of community and participation. A forum can be used also as part of the decision-making process within organizations.

An Interactive Discussion Forum infrastructure will be quite beneficial to a community of people, providing them with new means to communicate, exchange opinions and achieve their common goals.

A Forum is typically organized to discuss a certain issue. There is a moderator who invites a few experts who prepare a position statement and background material for the discussion. The forum involves an audience (either live or not) whose members can take part in the discussion. To take part in the discussion, each participant notifies his/her intention or directly transmits his/her contribution. To enable coordination of the discussion, organization of the contributions and keeping track of the decision process, the interventions must follow simple rules of dialectics: each participant must classify his intervention according to a limited set of categories:

Through the tools in the Interactive Discussion Forum (IDF) a participant can enter a Forum, by supplying his name, qualifications and a picture. The graphical interface of the IDF displays a panel with the issue under discussions, a board where the current speaker displays his material, which he comments by voice, a list of issues and a list of interventions. The list of interventions is structured according to the categories described above, so that one can get at a glance the state of the Forum. The moderator has the responsibility of selecting the next intervention to be multicast to the audience. The qualification of the contribution (person, qualification and category of intervention) are essential to make an appropriate choice that will speed converging on a decision. For instance it may be better to hear a counter-argument before an approval statement. In the list are present, with different color tags, the contributions that have already been heard and those still pending. Besides listening to the current speaker, a participant can listen to any other contribution, including previous ones, so that he can get up to date on parts of the discussion that he missed. Since the whole forum is archived, this allows asynchronous participation to a Forum. The list of interventions will be part of an argumentation structure which keeps track of the relation among the interventions, among the issues (for instance when an issue subsumes another or when it is composed of sub-issues).

The IDF are targeted to the Internet and World Wide Web. However the basic HTML model incorporated in current Web browser is driven from user interaction and does not provide the automatic update of the page as actions happen on the server, as in the case of events in a Forum. Therefore browsers must be extended with Java applets and plug-ins to distribute information and channels to each participant. Dynamic HTML and browser scripting can be used to support Java and plug-ins.

Figure 1. Sample of the interface for the IDF

To allow broader participation to Discussion Forum, we envisage setting up a wireless local area network (LAN) infrastructure. Wireless LAN connectivity allows users to roam freely throughout a facility or campus while maintaining high-speed wireless access to networked resources, including the Internet, World Wide Web, servers, printers and other users. Wireless LANs have gained strong popularity in a number of vertical markets, including the health-care, retail, manufacturing, warehousing, and academia. Today wireless LANs are becoming more widely recognized as a general-purpose connectivity alternative for a broad range of business customers.

Phillip Redman, mobile computing analyst at The Yankee Group in Boston, said: "Our research shows that 40% of a person's time in the office is spent away from his/her desk. Wireless local area networking is an important enabler, opening a new set of applications that require mobility and high-speed access to local networked resources". A sixfold expansion of the worldwide wireless LAN market is predicted by the year 2000.

Another subject of investigation will be the use of handheld computers, which are becoming quite popular, in creating mobile Networked Communities. The stability offered by Windows CE brought many new vendors into the handheld PC market and also helped increase awareness and sales of non-Windows CE-based handheld PCs. Handheld computing is one of the natural elements in a Wireless LAN and will provide new means to interact over the Net.

Hewlett-Packard Company is involved in experimenting Java on handheld computers. HP recently released a Java Virtual Machine optimized for handheld devices. Jonathan Roberts, group director of market development and Windows CE at Microsoft said: "HP's JVM is better optimized for even smaller embedded systems. HP's will run in about 500KB".

Timetable & milestones



1. Setup of wireless infrastructure

3rd month

2. prototype IDF server

6th month

3. prototype IDF clients

9th month

4. Experiments on Interactive Discussion Forum

12th month

Papers describing the results of the activity will be presented at appropriate international conferences, e.g. WebNet ’98 or WWW Conference ’99.

Activity Description

In order to achieve the project objectives, the following core technologies will be used:

        1. Multimedia, real-time communication applications over Mobile-IP networks
        2. Computer Supported Cooperative Work.

The project entails the development of:

        1. a suitable testbed infrastructure for Wireless LAN Connectivity. This infrastructure will be Mobile-IP based
        2. IDF server, stores and distributes over the Web the contents of each discussion forum
        3. graphical interfaces for Interactive Discussion Forum, one for the participants and one for the moderator

Wireless LAN Connectivity

The University of Pisa has an advanced private network based on extensive fiber optic cabling throughout the city, which reaches thousands of people in a few dozen of departments. This already existing infrastructure will be used as a static IP backbone for the testbed Wireless LAN infrastructure.

Two or three wireless access point will be installed in strategic locations to provide the widest coverage.

Dato and Bonito, who are the experts in radio transmission and multimedia protocols over IP, will carry out this task.

IDF server

An IDF server will be the machine to which all participants will connect to receive and submit their contributions. The machine will perform the services both of Web server, multimedia server and data base server for storing and distributing the discussion material. The service will be built in Java, using the servlet tools that our group has been developing.

Cisternino and Gullì will carry out this task.

Interfaces for the Interactive Discussion Forum

Two clients will be developed for the IDF: one for the moderator of the discussion, which enables him to order the discussion and give the right of speech to the participants; the second interface for the participants, which enable them to examine the argumentation structure, to book an intervention and to give their speech.

The content of a discussion will be stored according to the argumentation structure. Visual tools will be provided to navigate in the structure, selecting for instance on the kind of argument, on the proponents or opponents to an argument.

This task will be performed with the active participation of Attardi and Simi, and with design and programming support by Cisternino and Di Marco.


The participants are currently active in the following areas.


EUROsearch is an ESPRIT project which aims at building a pan-European federation of national search and categorization services. Initially comprising services from Italy, Spain, and Switzerland, the EUROsearch federation will constitute a distributed, multilingual searching service, permitting users to enter queries in their own, or their preferred language, and to carry out search and information retrieval over some or all of the federation’s national sites. Each national site will be responsible for maintaining and operating a search service dedicated to its own language, so that the needs of each language community will be catered for by native speakers of that language.

Our team is involved in the tasks of summarization and categorization of Web documents.

The Arianna project

SerRA, the networking center of the University of Pisa, participates in the Arianna project with Italia on Line and Interactive Telemedia Labs. Arianna is a specialized search engine for the italian Web space, which is currently the most complete as for number of pages and quantity of text. Arianna speaks italian, with an italian word stop list, an italian lexicon with italian stemming of words for the search and structure and content of its catalogue based on italian culture. The Arianna search robots are guided in the exploration of the Web space by a tool called testwww, which is part of a suite of tools to help the search designed and developed by the SerRA team, and made available as public domain software. The current Arianna architecture consists of two stages. The first stage performs the search by means of several robots gatherers which run in parallel on three hosts each one focusing its search on a different Autonomous System. He second stage performs a more refined selection of documents after they have been gathered from the net and then it performs an indexing of the document to be used for answering user queries. This stage is based on Fulcrum.

Arianna has received international awards such as The Best of Europe by Europe Online and by Beta.

The Arianna project is planning to extend its activity by establishing connections with similar search engines in other european countries to build a federated set of european search engines. Contacts are already being established with partners in Spain and Germany.

Arianna currently covers over 10,000 sites and a total of over 3 million Italian Web pages. The Arianna service is currently deployed by means of two servers located in Milan, and connected to the Infostrada network. However the growing number of queries submitted daily (over 200000) to these servers often lead to refusal of service. The HP grant allowed us to add a powerful server located in Pisa to perform gathering and indexing of material retrieved from the Italian Web. This allowed us to better balance the load among the servers, to reduce the duration of gathering and therefore to have shorter update cycles for the data base.


The University of Pisa has long term involvement in activities on networking and Internet in particular. The tradition in networking dates back to 1974 with the development of RPCNET by CNUCE, with the connection to the EARN network in 1984 and with the first italian connection to Internet in 1986.

At the University level, a special center for networking has been setup: SerRA (Servizi di Rete di Ateneo).

The center is responsible for:

The Frisco Project

FRISCO (Framework for Integrated Symbolic Numeric Computing is an ESPRIT LTR project which investigates and develops technologies which can be used to deliver highly efficient, versatile polynomial solvers to industrial users, in many areas of science and engineering: for example geometric modeling, robotics, chemical engineering, scheduling and electronics.

The project involves the following partners:

Universitad de Cantabria, Spain

INRIA Sophia Antipolis, France

IRMAR, France

The Numerical Algorithms Group Ltd, UK

Università di Pisa, Italy.

The Virtual Multiprocessor Project

The three topics, which bind this project together, are parallel and distributed systems, advanced compilation techniques and artificial intelligence. The research aim of this project is to demonstrate the advantages accruing from the combination of these three technologies in developing AI applications to run on distributed architectures and across local and wide-area networks. There are three identifiable strands: (i) the design, construction and application of reflective architectures; (ii) methodologies for the construction of knowledge bases; (iii) the construction of multi-agent systems. The network brings together 15 partners:

AI Lab, Vrije Universiteit Brussel, Belgium

GMD (FIT.KI) Bonn, Germany

GMD FIRST, Berlin, Germany

Christian Albrechts Universitaet Kiel, Germany

LSI, Universitat Politecnica de Catalunya, Spain

Universitat Rovira i Virgili, Tarragona, Spain

Institut d'Investigacio en Intelligencia Artificial at Campus Universitat Autonoma de

Barcelona, Bellaterra, Spain

INRIA, Rocquencourt, France

Ilog S.A., Gentilly, France

Dipartimento di Informatica, Università di Pisa, Italy

Istituto di Cibernetica, CNR, Naples, Italy

Dipartimento di Informatica ed Applicazioni, Università di Salerno, Italy

Mathematical Sciences, University of Bath, UK

Department of Electronics and Computer Science, University of Southampton, Italy

University of Warwick, UK

State of the Art

Group Communication

The tools for supporting group communication and group work can be distinguished between synchronous and asynchronous. For instance audio/video conferencing tools are synchronous while mailing lists or discussion groups are asynchronous.

In this project we will try to bring the liveness and interactivity of synchronous communications to discussion groups. In order to do so we need to overcome some technological limitations.

The basic client-server architecture of the Internet is primarily oriented towards user interactions via servers. Asynchronous tasks are well supported and there are a number of tools for supporting group discussions on the Web:

Web4Groups is a European initiative for setting up non-simultaneous group communication services for the WWW. The purpose is to satisfy the needs of users co-operating using telematic communication services.

WebNotes is a Web-based forums system which helps people gather and exchange ideas and information using the Web. WebNotes creates an organized, easy to navigate discussion forum. WebNotes is suited for on-line customer support, connecting widely scattered people within an organization, communicating effectively with business prospects and partners, and general discussion groups.

HyperNews supports conferencing and annotation on the web, includes modification, annotations may be links to other pages, multiple parents, moving of annotation trees, moderation, personal/group/public annotations, CGI interface, distributed servers.

ComMentor is a tool to share opinions in the Web developed at Stanford University.

For synchronous activities or applications which want to present information differently to different users, or where the information available changes frequently, the basic server model of the Web needs to be extended.

Audio/video conferencing is the most typical synchronous application and there are a number of conferencing tools on the market. Most of them exploit proprietary protocols and specific communication infrastructure. However the issues related to deploying such tools on Internet and on the Web are still to be solved.

Computer Supported Cooperative Work (CSCW) focuses mainly on the issues of supporting specific group activities. The problem in CSCW applications on the Web based is that they inherit a key property of the Web – that the protocol used (HTTP) is inherently stateless. If something changes in the application or in the server, the user remains unaware of this until he reloads the page (issues another HTTP request). As such, cooperative tools and applications that require any significant degree of synchronous interaction between the users, and/or the application, such as chatting or shared whiteboards, are not directly supported by the current Web architecture.

Most of the present Web-based CSCW systems are platform- and browser-dependent and all of them only address some specific form of synchronous interaction.

BSCW is an EC-funded project which is attempting to develop a general toolkit (MetaWeb) for supporting cooperative applications on the Web. However it is based on the idea of a "shared workspace'' which the members of a group establish for organising and coordinating their work.

COOPWeb s another EC-funded project focussing on common workspace provision and also supporting synchronous communication tools (audio, video). It is based on the shared workspace tool BSCW.

The idea of an argumentation framework is explored in project GEOMED, which aims at improving access to distributed heterogeneous Geographical Information Systems. The goal of the project is to make geographical information more accessible and affordable, so that public policy and planning procedures can be more transparent. GEOMED will integrate services for creating, disseminating and using geographical information in the sectors of urban planning, public administration, and environmental protection.

Handheld Computing

Industry's leading Handheld PC Vendors are converging to enable wireless LAN connectivity:

Proxim’s RangeLAN2 client adapters are offered as a one-piece PC Card and miniature modules for integration inside computing devices. RangeLAN2 technology is the foundation for the OpenAir wireless LAN standard adopted by the Wireless LAN Interoperability Forum, giving users interoperability with an open, multivendor wireless LAN. Proxim's RangeLAN2 products operate in the unlicensed 2.4 gigahertz (GHz) frequency band and utilize frequency hopping spread spectrum RF technology, which is both highly secure and exceptionally immune to interference. RangeLAN2 products operate at a data rate of 1.6 Mbps per channel with 15 available channels. This design can allow up to 24 Mbps aggregate bandwidth to support a growing user population. Proxim worked closely with Hewlett-Packard Company and Microsoft during the development of Windows CE driver software and compatibility testing with Proxim’s RangeLAN2 wireless LAN.

Hewlett-Packard Company is involved in handheld computing. Its top product HP 620LX Color Palmtop PC has a full support for Windows CE 2.0 and is ready to work in a Wireless environment using Proxim technology. It seems the natural platform to develop an Interactive Discussion Forum available to mobile users.

Internet Futures

Communication and sharing are the driving forces in the development of the Internet. The World Wide Web is an outgrowth of the need for sharing information, which is its main focus. Less attention has been given to new communication metaphors: the dominant mean of communication is still electronic mail (or newsgroups), while other approaches like push channels do not seem to gain wide user acceptance. Improvements in the Internet infrastructure will make transmission of multimedia material feasible, but this will not address the communication needs of the users. In a information society where attention is the most scarce resource, feeding even more material to people will only aggravate information overload. It is important to build intelligence in the tools that people use to communicate over the Internet. Such intelligence is for instance needed in helping user select appropriate material (in terms of relevance, quality and trustworthiness) or in interacting with other users.

One promising approach is to let such intelligence in the tools emerge directly from the interactions among intelligent people. We are exploring this approach in a tool for the categorization of Web material. As people organize the material in the Web into hypertexts, they incorporate their knowledge and understanding of the relations in the material. The Web then becomes storage of such collective knowledge. Our tool analyzes the structure of hypertexts and derives from it a categorization of documents that will facilitate retrieval.

We can apply similar idea to the communication interaction among users, identifying patterns of behavior which can be used in understanding needs, fulfilled or non-fulfilled expectations.

Rudimentary forms of intelligence can be achieved through interpreting the actions of users, but most likely language understanding will play an increasingly important role.

Publication of the proposals and results

All the software produced in the project will be placed in the Public Domain.

The group has already developed several Public Domain Software tools, among which we mention:

a suite of utilities for hierarchical exploration of the WWW, in particular a tool called testwww which establishes the presence of a Web server on a host and supplies statistical information on the server.

D. Dato, A. Gulli, G. Attardi,
Web hosts enumeration through DNS, WebNet 97.
Customisable Memory Management. The CMM is a memory management facility supporting complex memory intensive applications in C++. The CMM can manage several heaps, each one implementing a different storage discipline. The default heap for collectable objects uses the technique of mostly copying garbage collection, providing good performance and memory compaction. Other available heaps are a temporary heap, and a mark & sweep heap. CMM users can define their own heap classes for the specific storage requirements of their algorithms.
CMM has been used by
Sun Microsystem during the development of the language Oak, later become known as Java.
sources for CMM are available on FTP.
Papers on CMM:
G. Attardi and T. Flagella, A customisable memory management framework, Proc. of USENIX C++ Conference 1994, Cambridge, Massachusetts, April 1994, 123-142.
G. Attardi and T. Flagella, Customising object allocation, Object-Oriented Programming, ECOOP'94, LNCS n. 821, Springer-Verlag, Berlin, 1994, 320-343.
G. Attardi and T. Flagella, Memory Management in the PoSSo Solver, submitted for publication, 1994.
G. Attardi, T. Flagella and P. Iglio, Performance Tuning in a Customizable Collector, Memory Management, IWMM 95, LNCS n. 986, Springer-Verlag, Berlin, 179-196, 1995.
G. Attardi, T. Flagella and P. Iglio, A Customisable Memory Management Framework for C++, Software: Practice & Experience, 1998.
(Embeddable Common Lisp) is an implementation of Common Lisp designed for being embeddable into C based applications. ECL uses standard C calling conventions for Lisp compiled functions, which allows C programs to easily call Lisp functions and vice versa. No foreign function interface is required: data can be exchanged between C and Lisp with no need for conversion. ECL is based on a Common Runtime Support (CRS) which provides basic facilities for memory management, dynamic loading and dumping of binary images, support for multiple threads of execution. The CRS is built into a library that can be linked with the code of the application. ECL is modular: main modules are the program development tools (top level, debugger, trace, stepper), the compiler, and CLOS. A native implementation of CLOS is available in ECL: one can configure ECL with or without CLOS. A runtime version of ECL can be built with just the modules that are required by the application. The ECL compiler compiles from Lisp to C, and then invokes the GCC compiler to produce binaries.
sources for ECoLisp are available on FTP.
Polynomial System Solving is a computer algebra library for efficiently solving polynomial systems of equations. The library is suitable for incorporation into applications programs, in areas such as: kinematics, control theory, geometric modeling, and biochemistry. The library has been developed within the ESPRIT project Basic Research Action n. 6846 of the European Commission, in collaboration involving 16 European universities.

Java applications and applets

is a general tool for developing configuration applications running on the Web. Starting from a declarative description of the basic items to be chosen for the configuration and of the configuration constraints, the tool generates the HTML files for user guidance and the Java code for constraints checking. An interactive assistant for compiling and submitting plans of study has been built with the tool and deployed at our university.

G. Attardi, A.Cisternino, M. Simi, CompAss: A System for Plans of Study Compilation, AI*IA Conference 97, 1997.
G. Attardi, A. Cisternino, M. Simi, Building a Configuration Assistant on the Web, WebNet 97, 1997.
G. Attardi, A. Cisternino, M. Simi, Web-based Configuration Assistants, to appear in Artificial Intelligence for Engineering Design, Analysis and Manufacturing, 1998.
is a tool assisting in the preparation of Modello 740, the annual statement of earnings for the Italian Internal Revenue Service.
is a tool that allows starting several applications within a single Java Virtual Machine. It is useful for developing Java servlets that implement several services.
is a Java framework to develop servlets, handling multiple network connections with clients and keeping a log of activities.
Jamming.Net (Awards)
Jamming.Net is a Java server that distributes WWW traffic among a set of Web Server, in order to decrease load and latency. A graphic applet is provided to configure scheduling strategy, the number of servers and various fault prevention strategies. Jamming.Net also includes it local HTTP cache.
is a graphics client for the RoboCup Server. RoboMon displays graphically in real time a soccer match between teams of software agents.