Support Sense-making with Tools for Structuring a Conceptual Space

This abstract describes the design of a sense-making tool that assists users to structure their conceptual space and create a task report from the conceptual structure. Sense-making involves the ??? recognition of a knowledge gap, ??? seeking for information, ??? analyzing and synthesizing information to create an understanding, and possibly ??? producing a task output: a report, decision, or other type of output. A typical sense-making task is for an intelligence analyst to gather, analyze, and synthesize information related to some political figure or event, and to make recommendations for action. Everyday sense-making tasks, although less complex, share certain characteristics with the example. Many people use information systems as sense-making tools. Standard information retrieval systems can support reasonably well the search for pieces of relevant information when the user can identify her information need or knowledge gap to certain extent. However, in order to make sense of the information they found, users need to understand the pieces of interrelated information, identify patterns, and build on their previous knowledge to create an updated understanding. Most sense-making tasks are characterized by the interplay of searching for information and creating a structured representation of the situation, problem, or domain. In difficult sense-making tasks, sense makers often use external more or less structured representations to store the information for repeated manipulation and visualization (Stefik et al., 1999). Users need sense-making tools that facilitate the creation of such conceptual structures. Research in the areas of education (especially learning theories), cognitive psychology, and information extraction all bring useful insights into sense-making research. This study aims to integrate findings from these areas and build on that basis to design a sense-making tool that assists users in creating and using external representations for sense-making. Some questions to be explored are: 1. How do users build the concepts and relationships in their conceptual models? 2. How can the tool support the co-existence and transformation of different formats in which users represent their mental models of concepts and relationships? 3. How can automatic or computer-assisted extraction of entities and relationships assist users in building a representation? 4. Does the framework of topical relevance types assist users in making connections between retrieved information and conclusions to be reached and in identifying types of information to search for