13 Conclusion ................ 203

The postscript version of this chapter.

Table of Contents.

Chapter 13



The InfoCrystal (TM) is a powerful visual representation that uses a simple visual analogy to enable users to deal with some of the complexities involved in information retrieval. It is both a visualization tool and a visual query language. The InfoCrystal can visualize all the possible binary as well as continuous relationships among N concepts. In the binary case, it uses location, rank, shape, color and size coding to enable users to see in a single display how a large information space relates to their interests. In the continuous case, a novel polar representation has been presented that visualizes the relevance scores of the retrieved documents in a ranked order. The InfoCrystal represents all the possible Boolean queries involving its inputs in disjunctive normal form, which makes it very easy for users to modify a query. The InfoCrystal acts like a Boolean Calculator and users can use it to employ the expressive power of the Boolean retrieval approach and its broadening / narrowing techniques in a visual way. Users can assign relevance weights to the concepts of an InfoCrystal and formulate weighted queries by interacting with a threshold slider. The InfoCrystal has the added advantage that users can control in a visual way how to translate weighted queries into Boolean queries. Complex queries can be created by using the InfoCrystals as building blocks and organizing them in a hierarchical structure. Finally, the InfoCrystal has been generalized to visualize and formulate vector space queries. Hence, the InfoCrystal provides a visual framework that unifies the Exact and the Partial Matching approaches and enables users to take advantage of their respective strengths.

The results of a user study have been presented that compared the standard, text-based Boolean query language with the InfoCrystal in its most basic form. Subjects had to perform a recognition and generation task. The former asked users to recognize either the correct Boolean or InfoCrystal query from among three possible queries. The latter required subjects to generate a Boolean or InfoCrystal query that captured a given information need. These two tasks only tested a specific aspect of the InfoCrystal interface. Hence, this user study did not have the scope to fully evaluate the effectiveness of the complete InfoCrystal representation, but it produced the following useful results: 1) Novice users were able to successfully use the InfoCrystal, although they received only a short, fifteen minutes long tutorial. This second version of the tutorial made a big difference how well and quickly users could learn to use the InfoCrystal. Further improvements in the way novice users are instructed to use the InfoCrystal will help them to make full use of its rich set of features and the advantages that it has to offer. 2) The user study showed that the InfoCrystal, even at an early stage of development, performed as well as the familiar Boolean interface, although the study was biased in favor of the Boolean mode. 3) On the one hand, the study confirmed that the InfoCrystal is ideally suited for queries of the form "at most, exactly, or at least n out of m features". On the other hand, it showed that certain Boolean queries are more difficult to formulate using the InfoCrystal than the Boolean interface. However, we believe that users can improve their performance with more practice and if they have access to the enhancing features of the InfoCrystal that have been implemented, but were not made available during the experiments. 4) The user feedback concerning the InfoCrystal interface was very encouraging and it helped to pinpoint possible improvements. The user study shows that the InfoCrystal, even in its most basic form, can be successfully used by novice users and hence warrants further development.

This thesis has addressed the difficult problem of how to visualize information spaces that are abstract and do not have explicit spatial properties that can be exploited. The InfoCrystal provides a spatial overview of the data elements in an large information space and simultaneously provides visual cues about the content of the data elements. These opposing requirements have been resolved by exploiting the grouping principles used by the human visual system to make relationships between different, but related data elements visible and immediate. The InfoCrystal does not lock users into just one way of viewing the data. It helps them to decide how to proceed in the search process and how to control the output because the quantitative information associated with an interior icon. Further, the InfoCrystal is a visual representation that not only has descriptive power, because it enables users to see large amounts of information in a compact way, but that also has expressive power that enables users, for example, to interact with the data to issue retrieval commands. The InfoCrystal is a high-level retrieval interface because it encourages complex explorations and the creation of abstractions. It is flexible in terms of the retrieval methods and the data types used. The InfoCrystal supports "what-if" scenarios and it provides users with dynamic visual feedback.

The contribution of this thesis to the emerging field of Information Visualization consists of two parts. First, this thesis has demonstrated how information visualization offers ways to accomplish some of the needed improvements in information retrieval. It has provided a constructive proof that it is possible to visualize both Exact and Partial Matching methods in the same visual framework. Second, this thesis suggests directions for further research to develop the InfoCrystal into a tool that enables general users to make full use of its expressive power. The InfoCrystal has broad applications, because it offers a "visual machinery" to compare and relate any number of ordinary or fuzzy sets of arbitrary data items. It opens up new possibilities for complex data explorations. The InfoCrystal can be used as a general-purpose coordinator or generator of arbitrary data streams. It can be used as a Boolean Calculator in any domain where several criteria or requirements need to be coordinated. This thesis has briefly discussed how the InfoCrystal could be applied in domains such as: Internet Exploration, Document and Information Retrieval, Database Mining, Multimedia Editing. The Internet offers the opportunity to become the repository for diverse and rich bodies of knowledge. Hence, users need powerful visual tools that enable them to perform complex data explorations to discover meaningful new connections and opportunities. It is our hope that the InfoCrystal and its design principles represent a step in the desired direction. The overall goal of this thesis has been to contribute to the development of a Visual Retrieval Interface, where users can choose from a diverse set of visual tools to filter and visualize information.