Learning to Rank - Feature Engineering Using a Click Model

Voncina, Klemen (2018) Learning to Rank - Feature Engineering Using a Click Model. Bachelor's Thesis, Artificial Intelligence.

Preview	Text AI_BA_2018_KLEMENVONCINA.pdf Download (405kB) | Preview
	Text toestemming.pdf Restricted to Registered users only Download (94kB)

Abstract

Effective ranking in information retrieval is done, in part, by proper feature engineering. This paper explores a comparison between the functions of a click model and a ranking function in information retrieval. It then uses the output of a basic bi-class click model as a feature for training a ranking model. Training both of these different approaches on data from a commercial search engine we find that click model performance improves as the threshold for what is a click becomes more stringent and that using the output of a click model as a feature for ranking performs empirically worse than without this added feature.

Item Type:	Thesis (Bachelor's Thesis)
Supervisor name:	Wiering, M.A.
Degree programme:	Artificial Intelligence
Thesis type:	Bachelor's Thesis
Language:	English
Date Deposited:	25 Jul 2018
Last Modified:	27 Jul 2018 12:50
URI:	https://fse.studenttheses.ub.rug.nl/id/eprint/18052

Actions (login required)

View Item