ExpLOD: a Framework for Explaining Recommendations based on the Linked Open Data Cloud. Musto, C., Narducci, F., Lops, P., De Gemmis, M., & Semeraro, G. Proceedings of the 10th ACM Conference on Recommender Systems - RecSys '16, 2016.
ExpLOD: a Framework for Explaining Recommendations based on the Linked Open Data Cloud [link]Website  abstract   bibtex   
In this paper we present ExpLOD, a framework which ex-ploits the information available in the Linked Open Data (LOD) cloud to generate a natural language explanation of the suggestions produced by a recommendation algorithm. The methodology is based on building a graph in which the items liked by a user are connected to the items recom-mended through the properties available in the LOD cloud. Next, given this graph, we implemented some techniques to rank those properties and we used the most relevant ones to feed a module for generating explanations in natural lan-guage. In the experimental evaluation we performed a user study with 308 subjects aiming to investigate to what extent our explanation framework can lead to more transparent, trust-ful and engaging recommendations. The preliminary results provided us with encouraging findings, since our algorithm performed better than both a non-personalized explanation baseline and a popularity-based one.
@article{
 title = {ExpLOD: a Framework for Explaining Recommendations based on the Linked Open Data Cloud},
 type = {article},
 year = {2016},
 identifiers = {[object Object]},
 pages = {151-154},
 websites = {http://dl.acm.org/citation.cfm?doid=2959100.2959173},
 id = {429ab7cf-b58a-3549-b402-d73809569a7e},
 created = {2018-09-10T08:28:53.180Z},
 file_attached = {false},
 profile_id = {2ed0fe69-06a2-3e8b-9bc9-5bdb197f1120},
 group_id = {e795dbfa-5576-3499-9c01-6574f19bf7aa},
 last_modified = {2018-12-14T12:16:32.825Z},
 read = {true},
 starred = {false},
 authored = {false},
 confirmed = {true},
 hidden = {false},
 citation_key = {Musto2016},
 notes = {La evaluación del sistema consiste en mostrar diferentes estilos de explicación a los usuarios: basada en items (Item X is recommended, because you have tagged and rated items Y), basada en características (Item X is recommended, because it contains features a, b...), y tagcloud. Sólo con esta información, se pide al usuario puntuar la película. Se hace una comparación con las puntuaciones reales (obtenidas de puntuaciones reales de MovieLens) y con estos ratings basados en las explicaciones a través de la media y la desviación estándar.  <br/>La evaluación del framework se realizó con usuarios.<br/>1.- Se les pregunta sobre datos demográficos.<br/>2.- Se muestran explicaciones de diferentes estilos a los usuarios, entre ellas las generadas con este framework. Se pide evaluar la película según esto, Después se muestra el trailer al usuario y se pide evaluar la película otra vez.<br/>3.- Se pide al usuario que puntúe la explicación en base a los objetivos del sistema (transparencia, persuasión, satisfacción, confianza y efectividad).<br/>4.- Se comparan las diferentes puntuaciones antes y después de ver el trailer con medidas de evaluación (diferencia normalizada).},
 private_publication = {false},
 abstract = {In this paper we present ExpLOD, a framework which ex-ploits the information available in the Linked Open Data (LOD) cloud to generate a natural language explanation of the suggestions produced by a recommendation algorithm. The methodology is based on building a graph in which the items liked by a user are connected to the items recom-mended through the properties available in the LOD cloud. Next, given this graph, we implemented some techniques to rank those properties and we used the most relevant ones to feed a module for generating explanations in natural lan-guage. In the experimental evaluation we performed a user study with 308 subjects aiming to investigate to what extent our explanation framework can lead to more transparent, trust-ful and engaging recommendations. The preliminary results provided us with encouraging findings, since our algorithm performed better than both a non-personalized explanation baseline and a popularity-based one.},
 bibtype = {article},
 author = {Musto, Cataldo and Narducci, Fedelucio and Lops, Pasquale and De Gemmis, Marco and Semeraro, Giovanni},
 journal = {Proceedings of the 10th ACM Conference on Recommender Systems - RecSys '16}
}
Downloads: 0