Title :
Service Discovery Method Based on User Intent
Author :
Kataoka, Yasuyuki ; Watanabe, Toshio ; Tanaka, Kiyoshi ; Higashino, Suguru
Author_Institution :
NTT Service Evolution Labs., Nippon Telegraph & Telephonecorporation, Yokosuka, Japan
Abstract :
This paper introduces a method that supports a user who has only a vague idea of service use in discovering suitable mobile applications. The proposal allows the user to add application functions after selecting the purpose of service use from a list of novel application functions. The proposal is based on our concept of ISHI (Intent of Service and Human Interface) which works as an interface between service and user. ISHI can automatically determine the importance level and the novelty level of an application from various data. The proposed method is based on not only natural language but also the structured data of mobile applications. User experiments show that the proposed method yields, compared to the conventional method, better performance with respect to the time duration of application search and the frequency of recourse to new service functions.
Keywords :
human factors; information retrieval; mobile computing; natural languages; recommender systems; ISHI; application functions; application search; importance level; intent of service and human interface; mobile applications; natural language; novelty level; recourse frequency; service discovery method; structured data; time duration; user experiments; user intent; Androids; Humanoid robots; Indexing; Mobile communication; Natural languages; Silicon; Vectors; information recommendation; machine learning; service discovery; service indexing;
Conference_Titel :
Web Intelligence (WI) and Intelligent Agent Technologies (IAT), 2013 IEEE/WIC/ACM International Joint Conferences on
Conference_Location :
Atlanta, GA
Print_ISBN :
978-1-4799-2902-3
DOI :
10.1109/WI-IAT.2013.66