CD-SemMF: Cross-Domain Semantic Relatedness Based Matrix Factorization Model Enabled With Linked Open Data for User Cold Start Issue

Personalized recommendations to cold start user is one of the significant challenges in information filtering systems. Most existing systems inherited idea collaborative (CF) and avoids item metadata. For instance, consider a book domain whose metadata are author, publisher, language...etc. The elimination due diverse nature attributes cross-domain environment. Because this ideology, it hard for provide better users. Cold users people who new preferences unknown. A novel model proposed research work called Cross-Domain Semantic Relatedness based Matrix Factorization (CD-SemMF) resolves issue recommender system by exploiting Linked Open Data (LOD). “DBpedia” widely used LOD resource that contains semantic about different domains, which resolve above-said problem. Here, available connects items preferred target from various domains. fundamental knowledge graph links domains also benefits information. “LOD Semantic-Relatedness Measure” measure calculates closeness across relatedness measured instead similarity because diverse. Alternating Least Square method applied here learn preferences. provides relevant, personalized with gained source relatedness. Experimental evaluation done on Facebook Amazon datasets. It observed result CD-SemMF gives than baseline methods.