Providing Informational Support For Argumentation: The ISA Project Fahri Yetim German Dept. Of Information Systems Marmara University, Istanbul yetim@hisar.aenfor.marun.edu.tr yetim@hotmail.com and formal-structural presentation of discussion processes, as well as for extracting arguments [Conklin 1996], [Gordon and Karacapilidis 1999]. However, less attention had been paid to how Abstract informational support could be given to discussion processes, e.g. argumentation processes, in the course of This paper briefly presents the ISA project that which the supply of additional background information or addresses the issue of how argumentation facts might be appropriate. This requires a conceptual and processes can be supported by providing textual technical integration of information retrieval and information from document databases. The document management systems with web-based conceptual integration of data- and knowledge- discussion forums. Further technologies of a document- based technologies with discussion forums is oriented information and knowledge management have to illustrated, and the preliminary works for be developed and applied, which do not only contribute indexing documents as well as for providing to an ”organizational” knowledge base by representing, search mechanism are presented. extracting or distributing information from documents, but also make this information available during 1 Introduction discussions. Consider initiating a search for information in support of the current position, for example. As a consequence of global network technologies, human In this paper, the essential of the research project ISA is communication issues keep on moving into the center of presented. The first steps have been made in the field of research attention and add a new aspect to the previous information organization and the conception of an informational, presentational and transactional perspective, argumentation-oriented search for information. that is finding consensus concerning important issues from discourses [Kuhlen 1999]. The claim ”firms need to shift their attention from documents to discussions” 2 Related Work [Davenport and Prusak 1998, 106] emphasizes the Research on discussion processes has received growing importance of discourses for the practice of information interest in the Artificial Intelligence and Computer- and knowledge management. In that, since they facilitate supported Cooperative Work community during recent the organizing and recording of discussions, discussion years. Computer tools to facilitate discussion processes forums play a very particular part. Methodologically, they vary from simple classical tools (e-mail, mailing lists, are based on structured models of verbal argumentation. newsgroups, etc.) and web-based discussion forums, to Projects of research in this area have mainly concentrated more dedicated systems that meet a user’s wish to on how to use argumentation models for the archiving interpret and reason about knowledge during a discourse. For example, the system QuestMap [Conklin 1996] captures the key issues and ideas during meetings, and The copyright of this paper belongs to the paper’s authors. Permission to creates shared understanding within a knowledge team by copy without fee all or part of this material is granted provided that the placing messages, documents, and reference material copies are not made or distributed for direct commercial advantage. concerning a project onto the system’s ”whiteboard”, Proc. of the Third Int. Conf. on Practical Aspect of where interrelations are displayed graphically. A ”map” Knowledge Management (PAKM2000) then shows the line of argumentation that lead to key Basel, Switzerland, 30-31 Oct. 2000, (U. Reimer, ed.) decisions and plans. http://sunsite.informatic.rwth-aachen.de/Publications/CEUR-WS/Vol-34/ F. Yetim 23-1 Another category of systems does not only provide a supports hybrid (both frame- and rule-based) knowledge cognitive argumentation environment that monitors and representation methods. A database interface allows structures discussion processes, but also offers support for informational exchanges between the two sub-systems, for decision-making. For instance, the HERMES system example, the picking of explicit relations between [Karacapilidis and Papadias 1998] is intended to act as an information units by the KB system in order to deduce assistant who efficiently structures, and thus facilitates, implicit relations and their transcription to the DB system. communications. In particular, it acts as an advisor who recommends decisions by providing reasoning mechanisms. A system related to that is ZENO [Gordon Forum and Karacapilidis 1999]. The corresponding argumentation frameworks are variants of the informal IBIS model of argumentation [Rittel and Webber 1973]. These systems are related to the discussion forum element in the architecture of the ISA system. They monitor issues, positions, alternatives, preferences, etc., and refer them to each other. Most systems even provide a significant automation of the decision-making process. However, they do not address the issue of how the current process of argumentation could get further informational support. As Ballim and Karacapilidis [1998] pointed out, DB System KB System the following further tools of automation would be desirable: • an argument assistant that can follow and advice on the detailed content of an argument, and not just on its form; • an argument support tool that can peruse a document collection, in order to find relevant information units Figure 1: The ISA System that promote the agent’s assessment of a given argument. In the following sections, the components of these sub- Of course, a prerequisite to such tools would be the systems, and the applied methods are described in more capability of the computer system not only to understand detail. (at least partially) dialogs between human agents, being involved in decision-oriented argumentation processes, but also to assess the inherent structure and informational 4 Organization of Information Units content of documents. As a prototype tool, the ISA Providing informational support for discussions requires system is designed to represent documents in a way that an adequate method for the organization and the retrieval allows human agents to find the pieces of information of documents. Like a discourse, most (scientific) that are relevant to their current position in an documents are argumentative, containing a series of argumentation process. arguments that support or criticize a specific position. Therefore, an argumentation-based method for the 3 Architecture of the System indexing of documents - as proposed by Sillince [1992] in the context of information retrieval – seems appropriate In its architectural design, the ISA system integrates a for the support of discourses as well. forum where contributions to the discussion are handled, The design of the DB sub-system includes the manual with two sub-systems providing informational support indexing of documents and document units, facilitating (fig. 1): the search for informational support. There are the The data-based sub-system (DB system) serves as a stock following components: of argument-supporting information units (texts). For that, • Component for the construction of the vocabulary: (hyper)text bases are manually constructed, i.e. contents This allows the input of terms (concepts) and of of texts are described, and inter-textual relations are relations between terms (inter, intra as well as extra explicitly defined. The purpose of the knowledge-based linking). sub-system (KB system) is to deduce further implicit • Component for the construction of the text base: relations between information units, and in particular, to Documents are indexed formally by the name of the determine the (explicit and/or implicit) relations that are author(s), the title, etc., contents are described by relevant to the current position in an argumentation terms from the vocabulary and by argumentation process. relations (e.g., describes, criticizes, supports, etc.). Both sub-systems are still in the development stage. The The additional assignment of terms and relations to DB system is being developed under Visual-Basic and document sections allows an indexing of the inherent MS-Access. For the KB system GoldWorks III is being line of argumentation, for example, capturing the used, an expert system developing environment, which addressed problem or position, the solution to the F. Yetim 23-2 problem, the points of criticism or support of the 6 A Dialog Example position, etc. (e.g., ‘doc-1-section-1 describes information-management’, ‘doc-1-section-2 criticizes A dialog is intended to promote the decision on the car knowledge-management’). model that will be bought by an agent. There are various • Component for the definition of hyper-textual decision alternatives to discuss, e.g. the one of the relations: This allows to define term-document alternatives may be Mercedes, the other one BMW, etc. relations (e.g., ‘doc-1 describes information- For each alternative, there are pro and con arguments to management’) and document-document relations (e.g., take into account. ‘doc-1 criticizes doc-2’), both referring to documents as a whole (unlike the sectional references by the text Car model base component). These relations will be used by the KB sub-system for generating further relations. Alternative-1 Alternative-2 • Component for the integration of user models: Three general user types (student, lecturer, and librarian) Mercedes BMW have been considered with respect to differences in languages and presentation preferences. If informational support in favor of Mercedes is required, • Component for the search of texts in the text base: the following query may be raised: This allows the finding of documents or document sections by using search terms and argument patterns (Pro-argument Mercedes) (‘pro and contra’), or text connectors (e.g., ‘either ... or ...’). First, the KB system determines: (a) the explicit relations that belong to the group of 5 Knowledge-Based Support pro arguments; and (b) the combinations of explicit relations that have The knowledge-based support for retrieving relevant to be considered to find implicit relations documents to an argument is provided by the KB sub- involving pro arguments. system. This includes the determination of the explicit and implicit relations that are related to the current Then, the search for documents containing pro arguments argument type (pro or contra argument), and thus have to is performed on two levels. The system searches for: be taken into account for the searching the DB system. (1) document sections, whose descriptions include The method for determining explicit relations between the topic (Mercedes) and explicit relations of the texts is based on the modified version of the current argument type (pro argument); and argumentation grammar proposed by Sillince [1992], (2) documents, which are related to documents that where an argument is considered as a relation between describe the topic (Mercedes), whereas the terms X and Y (e.g., X criticizes Y). This grammar has corresponding (implicit or explicit) relation has been modified to the effect that relations are grouped into to be of the current argument type (pro argument) contra, pro, and neutral arguments, as shown in the following. The amount of information units that is found during this Argument: Term Pro-argument Term / search can be reduced in a further step by identifying Term Contra-argument Term/ similar argument patterns or rhetorical elements. For Term Neutral-argument Term example, documents could be considered as more relevant, Pro-argument: supports / … if they contain topics identified by the pattern ‘in this Contra-argument: criticizes / … paper’ and/or criticism identified by ‘however’, etc. Some Neutral-argument: mentions / … of these patterns are implemented in the present version of the DB sub-system as search options of a separate search In addition, logical rules have been implemented to component. enable the KB system to find out implicit relations between documents, which have not explicitly been 7 Conclusion defined in the DB system during the manual indexing process, but could hypothetically be assumed as valid. The ISA system thus far presented in this paper is still in Various types of rules are conceivable. Giving a simple the stage of development. The conception and example, the support relation between two documents X construction of the database (the DB sub-system) and and Y may be valid, if in X another document Z, and in some of the definition of logical rules for the deduction of Z the document Y is criticized. implicit relations (in the KB sub-system) is implemented. The technical integration of the DB and KB sub-systems Example: with the discussion forum has not been addressed yet. IF (?Dokument_X criticizes ?Dokument_Z) AND Further issues that remain to be addressed are the (?Dokument_Z critizeses ?Dokument_Y) following: adaptive visualization and ranking of search THEN results, integration of inference mechanisms for an (?Dokument_X supports ?Dokument_Y) automated identification of user preferences, extension of F. Yetim 23-3 search patterns to three languages (English, German, and Turkish). Finally, it is also important to address the practical issues, such as how long it takes to index a document in the level of detail needed, and whether the indexing process requires sophisticated personnel so that a partially automation of the process has to be considered. References Ballim, A.; Karacapilidis, N. (1998): Modeling Discourse Acts in Computer-Assisted Collaborative Decision Making. In: Reimer, U. (ed.): Proc. Of the 2 nd Int. Conference On Practical Aspects of Knowledge Management (PAKM98), 4.1 – 4.11. Conklin, J. (1996): Designing Organizational Memory: Preserving Intellectual Assets in a Knowledge Economy. GDDS Working Paper. Available at: http://www.gdss.com/DOM.htm Davenport, T.; Prusak, L. (1998): Working knowledge: how organizations manage what they know. Boston: Harvard Business School Press. Gordon, T.; Karacapilidis, N. (1999): The Zeno Argumentation Framework. In: Künstliche Intelligenz 3(1999),20-29. http://ais.gmd.de/MS/zeno/zenoSystem.html Karacapilidis, N.; Papadias, D. (1998): A Computational Approach for Argumentative Discourse in Multi- Agent Decision Making Environments. AI Communications Journal 11(1) 1998. Kuhlen, R. (1999): Die Konsequenzen von Informationsassistenten. Was bedeutet informationelle Autonomie oder wie kann Vertrauen in elektronische Dienste in offenen Informationsmärkten gesichert werden? Frankfurt a. Main: Suhrkamp. Rittel, H.; Webber, M (1973): Dilemmas in a General Theory of Planing. Policy Sciences, 155-169. Sillince, J. (1992): Argumentation-based indexing for information retrieval from learned articles. In: Journal of Documentation 4 (1992), 387-405. F. Yetim 23-4