Abstract : Discovery Hub is an exploratory search engine that helps users explore topics of interests for learning and leisure purposes.It makes use of a semantic spreading activation algorithm coupled with a sampling technique so that it does not require a preprocessing step. 1 Linked data based exploratory search and recommendation Exploratory search [2] systems are designed to assist users during expensive cognitive consuming search tasks such as learning or topic investigation. Theyprovide a high level of assistance during the navigation in the results space and advanced results explanations. In the past few years several works showed the interest of using linked datadatasets, and especially DBpedia 1 , for resources discovery in recommender and exploratory search systems. For instance Seevl 2 is a band recommender helping the discovery of musical content and artists on Youtube thanks to a recommendation algorithm and a faceted browsing functionality. MORE 3 is a movie recommender in the form of a Facebook application that performs film recommendations thanks to DBpedia.Aemoo 4 is an exploratory search system that offers a filtered view on the DBpediagraph and gives explanations on the relations between the resources shown to the user.Yovisto 5 is a video platform offeringan exploratory search featurethat proposes a ranked list of related topic besides search results.
https://hal.inria.fr/hal-01188678
Contributeur : Fabien Gandon <>
Soumis le : lundi 31 août 2015 - 13:45:00
Dernière modification le : mardi 8 septembre 2015 - 13:30:21
Document(s) archivé(s) le : mardi 1 décembre 2015 - 10:27:18
