Personalized Search

TL;DR

This paper introduces a personalized search re-ranking approach using the Obelix recommendation system, improving search result relevance in scientific databases like CERN's CDS through offline and online evaluations.

Contribution

It presents a novel re-ranking method with Obelix for personalized search, specifically applied to scientific publication databases, demonstrating improved relevance over traditional methods.

Findings

Personalized re-ranking outperforms latest first and word similarity methods.

Obelix effectively improves click position metrics.

Evaluation includes both offline and online user studies.

Abstract

As the volume of electronically available information grows, relevant items become harder to find. This work presents an approach to personalizing search results in scientific publication databases. This work focuses on re-ranking search results from existing search engines like Solr or ElasticSearch. This work also includes the development of Obelix, a new recommendation system used to re-rank search results. The project was proposed and performed at CERN, using the scientific publications available on the CERN Document Server (CDS). This work experiments with re-ranking using offline and online evaluation of users and documents in CDS. The experiments conclude that the personalized search result outperform both latest first and word similarity in terms of click position in the search result for global search in CDS.