Inferring Thematic Places from Spatially Referenced Natural Language Descriptions Benjamin Adams 1 and Grant McKenzie 2 1 Department of Computer Science, University of California, Santa Barbara, Santa Barbara, CA, USA 93106-5110, e-mail: badams@cs.ucsb.edu 2 Department of Geography, University of California, Santa Barbara, Santa Barbara, CA, USA 93106-4060 e-mail: grant.mckenzie@geog.ucsb.edu Abstract Places are more than just a location and spatial footprint. A sense of place is the result of subjective experience that a person has from be- ing in a place or from interacting with information about a place. Al- though it is difficult to directly model a person's conceptualization of sense of place in a computational representation, there exist many natural language data online that describe people’s experiences with places and which can be used to learn computational representations. In this paper we evaluate the usage of topic modeling on a set of travel blog entries to identify the themes that are most closely asso- ciated with places around the world. Using these representations we can calculate the similarity of places. In addition, by focusing on in- dividual or sets of topics we identify new regions where topics are most salient. Finally we discuss how temporal changes in sense of place can be evaluated using these methods. Keywords: Place, sense of place, topic modeling, text analysis, volunteered geographic in- formation. x.1 Introduction J. Nicholas Entrikin (Entrikin 1991) has written that narrative accountings of places are essential resources for understanding the world, because they provide “a distinct form of knowing that derives from the redescription of experience in terms of a synthesis of het- erogeneous phenomena.” A key aspect of these narratives is that they come from an individual point of view and therefore capture qualities of subjective experiences. Much volunteered geographic information (VGI) on the web comes in the form of unstructured,