Visual User Interface to Supporting Information Seeking Behavior in Context Searching

This paper proposes an interface to support information seeking behavior for context searching. Context searching is a search behavior based on trend-related queries (e.g., “What were the hot topics last year?”). Our proposed system facilitates such search behavior using trend-related queries. In the current search engine, users encounter difficulty in inputting a trend-related query using natural language because their query needs to have specific terms (e.g., iPod MAX @period). The users also experience difficulty in understanding the search results that are usually presented in a tabular form. To solve these issues, we present a new user interface to facilitate exploratory behavior and understanding. Initially, this interface identifies seven questions that we assume the users will use to perform a context search. Then, the system starts the search according to the classified questions, and the results are displayed in a timeline format.