Named Entity Recommendations to Enhance Multilingual Retrieval in Europeana.eu

Abstract

In the past years significant research efforts were invested towards the usage of Named Entity recommendation for improving information retrieval in large and heterogeneous data repositories. Such technology is employed nowadays to better understand user’s search intention, to improve search precision and to enhance user experience in web portals. Within the current paper we present a case study on recommending Named Entities for enhancing multilingual retrieval in Europe’s digital platform for cultural heritage. The challenges of designing an entity auto-suggestion service able to effectively support users searching for information in Europeana.eu are described in the paper together with a preliminary experimental evaluation and the outline indicating the directions for future development.