Summarizing and Visualizing Web People Search Results

People search is one major search activity on the Web. If the list of people search results is merely “person 1, person 2, . . . and so on,” users have difficulty determining which person clusters they should select. In this paper, we present a project that summarizes and visualizes Web people search results to help users select person clusters more easily. We explore three ways of summarizing people: (a) selecting terms from the extracted information, (b) combining the extracted information, and (c) obtaining information from external databases referring to the extracted information. To visualize people, we present three types of interfaces: (a) tables, (b) two-dimensional space, and (c) map interfaces. We report the two results of the project. (1) We investigated algorithms for distinguishing individuals with identical names and three ways of summarizing people: extracting keywords, prefectures and vocations; combining vocation-related information; and obtaining locations. (2) We developed prototypes to display separated individuals by three types of interfaces.