Extraction of Cognition Results of Travel Routes with a Thesaurus

Extraction of Cognition Results of Travel Routes with a Thesaurus Kazutaka TAKAO Yasuo ASAKURA Department of Global Development Science, Dept. of Science of Regional and Built Env., Graduate School of Science and Technology, Graduate School of Science and Technology, Kobe University Kobe University 1-1 Rokkodai-cho, Nada-ku, Kobe 1-1 Rokkodai-cho, Nada-ku, Kobe 657-8501, Japan 657-8501, Japan 003d912n@y02.kobe-u.ac.jp asakura@kobe-u.ac.jp Abstract We are attempting to model travel route choice behaviour with language to describe the thinking process of travelers because words can directly and clearly reflect their psychological states from a bottom-up viewpoint. This paper shows a method that extracts impressions and feelings, i.e., cognition results of travel routes, out of open-ended questionnaire texts with a thesaurus. Complex words are also allowed as cognition results. Additional considerations and training contents are also reported. Finally, an experiment on the extraction of cognition results from unseen texts is reported. 1 Introduction 1.1 Background We are attempting to linguistically model spatial cognition and travel route choice behaviour. In the field of travel engineering, a unit of travel called a "trip" is expressed as movement from an origin to a destination; route choice behaviour is expressed as choice from the set of alternatives in each trip. Existing city and traffic infrastructure planning are mainly framed according to economical effects and demand estimations. Many studies have approached route choice behaviour from such top-down standpoints. These studies are more interested in the results of travel behaviour rather than the psychological state. Many have used numerical equations to explain behaviour quantitatively, and psychological factors are often expressed with internal variables. On the other hand, when developing new products, a company must analyze customers' awareness from a bottom-up viewpoint. In the same way, it is also important for traffic infrastructure planning to analyze from a bottom-up viewpoint, observing travelers' psychological states when making choices. We expect to handle psychological factors more clearly with words. Psychological states of route choice behaviour can be explained by words with analyzing open-ended questionnaire texts. 1.2 Modeling of Travel Route Choice Behaviour Route choice behaviour "recognizes" the characteristics of the alternatives (i.e. travel routes) in a choice set and "chooses" one route. Therefore, route choice behaviour can be expressed as a two-stage process that places the "cognition result" between the travel route and the behaviour as shown in Figure 1. The first stage recognizes the characteristics, feelings, and impressions of each route in the PACLIC 18, December 8th-10th, 2004, Waseda University, Tokyo - 259 - choice set, i.e., cognition results. The second stage chooses a route by evaluating cognition results. These stages are analyzed with open-ended questionnaire texts. This paper describes the first stage, i.e., the "recognize" stage whose objective is to extract enough cognition results to handle the travel route choosing process in the second stage. This paper introduces a process that uses a thesaurus to extract cognition results from open-ended questionnaire texts. Figure 1: Route Choice Process 1.3 Aspect in EBA Model Extracted cognition results are handled in the second stage. Therefore, we have …

Content truncated to 500 words