Using Complex Tasks and User Reviews to Help Users Reach Search Goals on Smartphone

• Main Authors: Shih-YangKo, 柯世洋
• Other Authors: Wen-Hsiang Lu
• Format: Others
• Language: en_US
• Published: 2016
• Online Access: http://ndltd.ncl.edu.tw/handle/82351922612703447395

碩士 === 國立成功大學 === 資訊工程學系 === 104 === In recent years, smartphones have become more and more popular. Users thus heavily rely on smartphones to get various information. According to the research in 2014, apps usage grew 76% than previous year. Apps have become an indispensable tool day by day. Conventional search engines such like Google Play usually consider a search query as a simple task. However, the queries are often driven by complex tasks. The results are often less than expected. In this work, we utilize the structure of complex task query and subtask goals and help users to find apps and location entities. Our Complex-Task-based Service Recommendation Model (CTSRM) contains three stages. Generating the subtask goals from complex task query, identifying the subtasks that could do on smartphones, and recommending suitable apps and location entities based on user review opinions for every subtask. CTSRM can recommend apps including subtask goals. For those subtasks which cannot do on smartphones, we recommend location entities to assist user accomplish their complex tasks.