Towards Culturally-Situated Agent Which Can Detect Cultural Differences Heeryon Cho 1 , Naomi Yamashita 2 , and Toru Ishida 1 1 Department of Social Informatics, Kyoto University, Kyoto 606-8501, Japan cho@ai.soc.i.kyoto-u.ac.jp, ishida@i.kyoto-u.ac.jp 2 Media Interaction Principle Open Laboratory, NTT Communication Science Laboratories, Kyoto 619-0237, Japan naomi@cslab.kecl.ntt.co.jp Abstract. A method to calculate the semantic dissimilarity in two coun- tries’ pictogram interpretations is proposed. Two countries’ pictogram interpretation words are mapped to SUMO classes via WordNet2SUMO. Appropriate concept weights are assigned to SUMO classes using the in- terpretation ratios. The edges between the two SUMO classes are counted to obtain the path length of the two classes. Three bipartite graphs are generated using the classes and edges to calculate the between-country vs. within-country dissimilarity in pictogram interpretations. Preliminary result showed that human assessment of interpretation dissimilarity does not always correspond to concept-level dissimilarity in the ontology. Keywords: interpretation, cultural difference, detection, ontology. 1 Introduction Our goal is to build an agent which can automatically detect cultural differences. Existing literatures on culturally-situated agents have tackled the problem of cooperation between agents with different cultural backgrounds[1] or the problem of bridging humans with different cultural backgrounds[2]. The former focuses on conflict resolution while the latter focuses on mediation. In this paper, we tackle the problem of automatically detecting cultural differences based on human- provided interpretations. We use pictogram as a symbolic medium to collect human interpretations from two different cultures. Pictograms have clear pictorial similarities with some object[3], and one who can recognize the object depicted in the pictogram can interpret the meaning associated with the object. Pictorial symbols, however, are not universally in- terpretable. For example, the cow is a source of nourishment to westerners who drink milk and eat its meat, but it is an object of veneration to many people in India. Hence, a picture of a cow could be interpreted quite differently by Protes- tants and Hindus[4]. We conducted a human cultural assessment experiment using U.S.–Japan pictogram interpretations as stimulus. Experimental findings revealed that human subjects looked at similarities and differences in two coun- tries’ interpretations when assessing cultural differences. Based on this finding, A. Ghose, G. Governatori, and R. Sadananda (Eds.): PRIMA 2007, LNAI 5044, pp. 458–463, 2009. c  Springer-Verlag Berlin Heidelberg 2009