EvalPCIR - Evaluation of Personalised Conversational Information Retrieval

In the last few years the research area of Personalised Information Retrieval (PIR) has been receiving a lot of attention. PIR aims at improving the retrieval process by taking into account the particular characteristics and needs of individual users. In this proposal we are concerned with the latest area of research in IR which ideals with the use of conversational interfaces to better understand the user information need. This research area is known as Conversational Information Retrieval (CIR). In particular, we plan to start the design of a new approach for the evaluation of Personalised CIR (PCIR) that will properly take into account the individual (or at the very least categorical) differences between users in assessing the effectiveness of the retrieval process and of the conversational interaction. Prof Kando is a leading scientists in the evaluation of PIR and during the visit we will design of a new evaluation methodology for PCIR. The results of this effort will be very valuable to the entire IR community and possibly be applied in the context of NTCIR or CLEF.