Context-Specific grounding of web natural descriptions to human-centered situations
作者：
Highlights：
• This novel W2S method is effective in grounding web natural descriptions into human-centered situations. With the grounded knowledge, human-centered situations could be effectively understood.
• The influence of environmental context towards humans’ activities could be measured. The context involvements in situation representation could improve the accuracy of situation understanding.
• The proposed semantic analysis method could evaluate the situation-involvement degree of the knowledge entities (i.e., objects and environmental context) and their relations. Based on the involvement degree, the key features including objects, environmental context, and their mutual relations could be extracted to accurately represent a situation.
摘要
•This novel W2S method is effective in grounding web natural descriptions into human-centered situations. With the grounded knowledge, human-centered situations could be effectively understood.•The influence of environmental context towards humans’ activities could be measured. The context involvements in situation representation could improve the accuracy of situation understanding.•The proposed semantic analysis method could evaluate the situation-involvement degree of the knowledge entities (i.e., objects and environmental context) and their relations. Based on the involvement degree, the key features including objects, environmental context, and their mutual relations could be extracted to accurately represent a situation.
论文关键词：Assistive robot,Decision making,Human-centered situation,Web information retrieval,Environmental context
论文评审过程：Received 4 March 2016, Revised 25 July 2016, Accepted 29 July 2016, Available online 18 August 2016, Version of Record 23 September 2016.
论文官网地址：https://doi.org/10.1016/j.knosys.2016.07.037
