Exploiting Semantic Template for Message Summarization for Mobile Devices

Abstract

In contrast with increasing amount of information on the web, mobile users are suffering from low hardware capacity, poor interface, and high communication cost problems for using their wireless devices. In this paper, we propose a text summarization-based framework for supporting users in wireless network environment. Thereby, we focus on ontology-based template generation. This system can extract and send particular parts of a text which are relevant to the corresponding user's contexts, instead of sending the full text itself. The templates can be generated by not only user's manual input but also semantic tagging, which is a process of categorizing keywords into the most relevant concepts on the ontologies. Hence, these semantic templates can be applied as a set of rules highlighting (i.e., discriminating) the specific parts of a given message. For evaluating the proposed approach, we have conducted experiments on wireless reverse auction environment in which participants can instantly send and receive the bidding messages through their mobile devices.