Search Engine That Listens (SETL)
Funded by 12th Japan CORE projects (CORE-12), Microsoft Research Asia
SETL was selected as one of the best MSR CORE Projects in 2016/17!
Searching is often a means to an underlying larger task called work task. For example, to decide a venue for a conference, one might need to gather various kinds of information about local hotels. However, existing search engines leave the burden of search task to end-users who must recognise information needs and formulate queries. This project aims to reduce the amount of effort people put to search tasks, allowing them to spend more resource for work tasks. To achieve this, a framework called SETL (Search Engine That Listens) is proposed to exploit conversations, that naturally occur during collaborative work, as a source of information needs.
- Sosuke Shiga, Hideo Joho, Roi Blanco, Johanne Trippas, and Mark Sanderson: Modelling Information Needs in Collaborative Search Conversations. SIGIR 2017: to appear.