Workshop Report

William Turner
LIMSI-CNRS
E-mail: William.Turner@limsi.fr

Francis I. Njilié Yap
ENA Presence Manager
E-mail: presencemanager@ethnonet-africa.org

Within the framework of Ethno-Net Africa (ENA) - a project of the Management of Social Transformations Programme (MOST) of the United Nations Educational, Scientific and Cultural Organization (UNESCO) a seminar was held from 18-21 March 2002 in Yaounde - Cameroon on "New Information and Communication Technologies (ICT's) and their use in training and co-operation in scientific research". Jointly organized by the "International Centre for Applied Social Science Research and Training" (ICASSRT), UNESCO and the "Laboratoire d'Informatique pour la Mécanique et les Sciences de l'Ingénieur" (LIMSI) of the "Centre national de la recherche scientifique" (CNRS), France, the seminar brought together twelve of participants from five African countries (Benin, Cameroon, Congo-Brazzaville, Côte d'Ivoire, Nigeria) and France.
In his introductory statement, the UNESCO representative Mr Halward Plinkert portrayed UNESCO's strategic objective to enhance access to more diverse information sources and thus to facilitate comparable international scientific research in order to enhance learning opportunities through access to diversified contents and delivery systems and to strengthen capacities for scientific research, information sharing, and to promote the use of ICTs for capacity building. In the framework of this objective, UNESCO/MOST contracted the research team of Prof. Turner to develop a software: the "Information Support System" (ISS).

The ISS permanently scans online news agencies, newspapers and an extendable list of other Internet information sources such as the observatories of Amnesty International and Human Rights Watch. It regroups information from those sources with the use of a "natural language analysis" in text format. The added value of this software in comparison to existing Internet search-engines is first, the control of the information sources, second, a natural language analysis of the articles, recognising related issues monitored by other researchers, and third, the feature that all articles can be commented on by the users of the ISS. Those comments are stored together with the articles in a newly established database, accessible through the recently decentralised website of the Ethno-Net Africa Project. Mr Plinkert suggested that this tool should be evaluated in the light of the following question: "What are the possibilities and the limitations of a software application for international scientific co-operation; In how far can it contribute to the necessary improvements concerning the mandate of Ethno-Net Africa and can it help to refocus its activities on policy relevant social science research; Is it a useful application to meet the demands of the members of the Ethno-Net Africa to improve their still insufficient access to the Internet based information sources, and their limitations to co-operate internationally and hence the relatively low efficiency regarding the creation of new, policy relevant knowledge. However, the members of the Workshop did not want this question to blurr the difficulties met by the National Monitoring Units in their functioning. The lack of technical and logistic infrastructures in the various countries was unanimously considered the main obstacle in their activities.

Mr Plinkert replied that even better infrastructures, would not answer the main problem: the relevance content for scientific analysis and the importance to differentiate between access to information and the production of new knowledge. The Workshop was thus organized to train the researchers of ENA in the features of the ISS relating to its adaptability, its new modes of co-operation and communication (the evolution from end-user to producer, editor, reader and distributor at the same time) and its possible function as a peer review mechanism to continuously monitor the scientific quality of ENA reports.

After the discussion on the Workshop objectives, the national coordinators presented on one hand, an analysis of the situation of the ethnic conflicts in their respective countries during the years 2000 and 2001 and on the other hand, the report of activities of their National Monitoring Unit providing an analysis of the geopolitics of the ethnic conflicts recurred in each of the regions. The ethnic conflicts appear to be strongly connected with issues such as democracy, elections, political debate, water, land, economy, etc… Although every country has its specificities, it was obvious that ethnicity is instrumentalized in many cases in favor of certain groups.

The National Monitoring Units identified the centralized coordination of the project, the mobilization of the members at national level, the lacking information infrastructure and the irregularity of feedback from the project management as main obstacles to good functioning.

The central subject of the seminar was presented by the representative of the LIMSI/CNRS, Mr. William Turner. He spoke on the transformation of information into knowledge for collective action relating to the ISS.

The first interventions made clear that this transformation passes by the access, the socio-cognitive evaluation, the interpretation and the distribution of the information.

The ISS is designed to support both the "in-house" and "co-operative" research activity of the ENA National Monitoring Units, and help organize a "science watch" for systematically drawing attention to new, emerging subjects of interest for social science research and analysis. It is still in the process of finalization to implement the useful and important comments by ENA members and by LIMSI-CNRS to improve the monitoring of the mass of information on ethnic conflicts and social change in Africa which is available on Web Sites of electronic newspapers and a wide variety of other American, European and African sources.

The ENA Data Archive will be constantly updated by periodically downloading Web-based information resources located at specific Internet sites. With each update, new documents that are relevant to an specific profile have to be identified and extracted for delivery to the appropriate country in Africa. This "filtering" task is based on expressions which might be used to index the article or might be appropriate for considering the event as being relevant for analyzing ethnic conflicts. Any Ethno-Net user can define his set of expressions (profile), which are pertinent according to his research area. A group of documents originating from a particular source is called corpus.

Example:

Profiles in French

PERSON A: Litiges fonciers -- Conflits halieutiques -- Régions

PERSON B: Quotas ethniques -- Obscurantisme politique -- Exclusion
Bamileké

PERSON C: Démocratie béninoise -- Régionalisme -- Ethnocentrisme
Vie politique et conflits politiques

PERSON D: Partis politiques -- Micro tribalisme -- Ninjas -- Milice -- Conflits ethniques

Profiles in English

PERSON E: Ethnic conflicts -- Sharia -- Niger Delta -- Marginalisation

PERSON F: UNESCO -- MOST -- Ethnic conflicts -- Decision Makers

The profiles are posted to the system's administrator via the Internet, who carries out the technical tasks of filtering the information sources designated by the user. The filtering is done overnight and, subsequently, the day after posting a filtering strategy (profile + information source), the user can open his Internet browser to see the results through the ISS interface.

The user has the choice between three options which we will explain focusing on concrete examples practiced during the workshop.

Option 1: Information relative to the specific subject area of an user (Deepening)

Assume Mr. A is the current user. After he activated his profile, he had access to the list of paragraphs that contain the expressions of his profile. In other words, the system will provide him with information that is directly related to his "in-house" research, his specific area of interest. A link for accessing the document that contains the detected paragraph is available near the paragraph.

Option 2: Information that broadens the scope of the subject area (Widening)

Now, Mr. B is operating. When he activates option 2 within his profile, he then has access to the list of paragraphs that contain words in the expressions of his profile that are used by other members of ENA. For example, his profile contains the expression "conflits halieutiques" whereas Mr. C's profile contains the expression "conflits éthniques". Both expressions contain the word "conflicts"; for Mr. B ethnic conflicts often erupt e.g. in the Ivory Coasts around water issues, and he prefers using the expression "conflits halieutiques" which is specific to his country context; whereas Mr. C chose to use the more general term "ethnic conflicts" in the Congo context. However, the system will alert both Mr. B and Mr. C to the fact that they are both interested in situations of open conflict.

Option 3: A list of words for up-dating your subject area profile (Emergence)

The ISS offers help in updating subject area profiles by automatically extracting from the source material (French corpus and English corpus), the 100 expressions which characterize the corpus.

Generally, a corpus is built using a single point of view to define a document collection, and this is when science watch comes into play. In the English corpus, two completely different points of view were combined (artificially) to produce a corpus for experimenting with the functions of the ISS system. The French corpus is much more coherent. The papers presented by Joachim Agbroffi, Christian Agossou, Norbert Gami, Dieudonné Zognong concern ethnic conflicts in their different countries. The terms that appear on the table of extracted expressions are all related to this general theme. Consequently, when theyconsult the list, they can immediately identify the terms specific to their own papers and those extracted from other papers, and this is the basis of a science watch.

For example, we saw that Mr A has an interest for the general question of "obscurantisme politique". Mr B filed a paper in which he speaks of the "mécontentement des Ivoiriens face aux étrangers" which the machine has extracted from the French corpus as being a potentially useful term for science watch. Science watch aims at offering a list of terms for visualizing links between concepts that structure a corpus: for example, we might expect that one of the reasons why foreigners are sometimes considered with distrust and hard feeling is a context of political obscurantism; the fact that "obscurantisme politique" and "mécontentement face aux étrangers" appear on the same list could prompt either Mr A or Mr B to explore this link, and even to up-date their respective profiles. The intellectual proximity of the expressions on the list is justified by the fact that they are extracted from a coherent set of documents addressing the general theme of ethnic conflicts.

The following two tables provide examples of the expressions extracted from the corpuses English corpus and French corpus.

A sample of expressions automatically extracted from the English corpus

A sample of expressions automatically extracted from the French corpus

Cooperative research requires discussion and interaction with the other members of ENA. The system ISS uses a forum to allow this discussion and interaction.

In the example, the system assumes that because Mr A is interested in the subject "obscurantisme politique", other members of ENA might want to share with him their point of view on the question. In order to do so, they type their password credentials and comments in the appropriate fields of the dialogue box and this information will appear on the bulletin board of the ISS so that other members of the Network can also interact.

Conceptually, the results supplied by the system can afterward be exploited by the members of the network ENA to, among other things, produce objective analysis, extract data for its database (Web), prepare colloquiums, lead debates (Forum), etc. A stage of validation of the various productions which ensue from it will then be necessary before scattering wide broadcasting, in the intention of others, the decision-makers.

The main recommendation stemming from the seminar concerns the follow-up of this project and the application, in the short-term of the ISS. The participants also formulated very useful suggestions for further product improvements to make the ISS easier and more efficient to use.

Conclusion
The Workshop was a success in the sense that participants gained a critical perspective on how their research practices might evolve through adoption of new information and communication technologies. That agreed that much remains to be done in order to develop a technical as well as anorganizational framework to meet the needs of ENA.

The first point concerns the information infrastructure in most African countries. National and international investment programmes to develop infrastructures are underway and so one can expect that Africa's current handicaps in this field will hopefully be overcomed in the future, but there is still a long way to go. For example, even the ENA system administrator in Yaoundé has only limited access to Internet at extremely high costs. Solutions to this problem as to offer remote distance access to LIMSI/CNRS computers, or to use satellite broadcasting technologies have to be discussed.

The second point concerns the efforts of UNESCO to produce high quality social science research in order to provide decision-makers with knowledge for successfully managing social transformations. We saw in Yaoundé that the knowledge produced by the ENA network will necessarily be the product of a wide variety of in-house research programs carried out in different contexts, using different theories and different observation protocols. Up to now, the project coordinator did not identify procedures to reach agreement on how to aggregate across the heterogeneous research results produced by ENA's participating members. The Information Support System is designed to carry out an in-depth analysis of a subject area, to promote cooperative research and to monitor available information sources to detect new, emerging subjects of interest. Although members of the Workshop recognized the necessity of building an infrastructure to support these procedures, it is as yet too early to understand their value for building confidence in the coherence of a distributed scientific practice.

© The ideas and opinions expressed in this article are those of the author
and do not necessarily reflect the views of UNESCO.

© Les idées et opinions exprimées dans cet article sont celles de l'auteur
et n’engagent pas la responsabilité de l´UNESCO.