A survey of user behaviour models for search results ranking

S. I. Nikolenko^ab, A. A. Fishkov<sup>c

a</sup> St. Petersburg Department of V. A. Steklov Institute of Mathematics, Russian Academy of Sciences
^b St. Petersburg Academic University — Nanotechnology Research and Education Centre of the Russian Academy of Sciences (the Academic University)
^c Saint-Petersburg State Polytechnical University

Abstract: User behaviour models represent important features used to enhance web search results; they are usually probabilistic models learned from logs of user actions (click logs). We present a survey of modern user behaviour models and review how behaviour models are combined with other features in the ranking function.

Keywords: user behaviour models, probabilistic graphical models, boosting.

UDC: 004.8

Received: 18.06.2012