资 源 简 介
Instant messaging (IM) is a popular form of real-time remote communication between people via a network such as the Internet. However, current IM applications still lack intelligence to provide a natural approach for human-computer interaction. For example, an IM user still needs to switch among multiple conversation windows when talking to several people simultaneously. Obviously, this brings inconvenience to IM users in practice. But few studies have been conducted to overcome this issue. In this project, we propose a novel approach to auto-detect receiver of each message in multiple simultaneous IM conversations by applying a semantic relevance model that takes both messages contexts and contents into account. Thus, applying our approach could greatly improve the usability of current IM applications in order to support multiple simultaneous two-party conversations.