Smith Search: Opinion-Based Restaurant Search Engine

Jaehoon Choi, Donghyeon Kim, Donghee Choi, Sangrak Lim, Seongsoon Kim, Jaewoo Kang, Youngjae Choi

Abstract: Search engines have become an important decision-making tool today. Unfortunately, they still need to improve in answering complex queries. The answers to complex decision-making queries such as ``best burgers and fries'' and ``good restaurants for anniversary dinner,'' are often subjective. The most relevant answer to the query can be obtained by only collecting people's opinions about the query, which are expressed in various venues on the Web. Collected opinions are converted into a ``consensus'' list. All of this should be processed at query time, which is impossible under the current search paradigm. To address this problem, we introduce Smith, a novel opinion-based restaurant search engine. Smith actively processes opinions on the Web, blogs, review boards, and other forms of social media at index time, and produces consensus answers from opinions at query time. The Smith search app (iOS) is available for download at http://www.smithsearches.com/introduction/.