Query Recommendation to Draw a Laugh from Web Searchers

This study proposes a system, which shows funny term pairs when searchers issue a query into Web search engines. The proposed system analyzes the following two factors for a term pair in a given query: the unexpectedness and the semantic conflict. The experimental result showed that the proposed method provided a larger number of funny term pairs for queries than the baseline methods. Although the proposed method was not the best based on the average value, it can still offer opportunities for searchers to laugh and feel cheery when issuing queries into the Web search engines.