Learning to Identify Complementary Products from DBpedia

Identifying the complementary relationship between products, like a cartridge to a printer, is a very useful technique to provide recommendations. These are typically purchased together or within a short time frame and thus online retailers benefit from it. Existing approaches rely heavily on transactions and therefore they suffer from: (1) the cold start problem for new products; (2) the inability to produce good results for infrequently bought products; (3) the inability to explain why two products are complementary. We propose a framework that aims at alleviating these problems by exploiting a knowledge graph (DBpedia) in addition to products’ available information such as titles, descriptions and categories rather than transactions. Our problem is modeled as a classification task on a set of product pairs. Our starting point is the semantic paths in the knowledge graph linking between product attributes, from which we model product features. Then, having a labeled set of product pairs we learn a model; and finally, we use this model to predict complementary products for an unseen set of products. Our experiments on a real world dataset from Amazon show high performance of our framework in predicting whether one product is complementary to another one.

[1]  Iván Cantador,et al.  A generic semantic-based framework for cross-domain recommendation , 2011, HetRec '11.

[2]  Marcelo Arenas,et al.  Semantics and complexity of SPARQL , 2006, TODS.

[3]  Bracha Shapira,et al.  Recommender Systems Handbook , 2015, Springer US.

[4]  Xavier Serra,et al.  A Semantic Hybrid Approach for Sound Recommendation , 2015, WWW.

[5]  Yang Song,et al.  Task Trail: An Effective Segmentation of User Search Behavior , 2014, IEEE Transactions on Knowledge and Data Engineering.

[6]  Bo Yang,et al.  Graph-based features for supervised link prediction , 2011, The 2011 International Joint Conference on Neural Networks.

[7]  Emine Yilmaz,et al.  Uncovering Task Based Behavioral Heterogeneities in Online Search Behavior , 2016, SIGIR.

[8]  I. C. Mogotsi,et al.  Christopher D. Manning, Prabhakar Raghavan, and Hinrich Schütze: Introduction to information retrieval , 2010, Information Retrieval.

[9]  Nitesh V. Chawla,et al.  Market basket analysis with networks , 2011, Social Network Analysis and Mining.

[10]  Hans-Peter Kriegel,et al.  Factorizing YAGO: scalable machine learning for linked data , 2012, WWW.

[11]  Hongbo Deng,et al.  Identifying and labeling search tasks via query-based hawkes processes , 2014, KDD.

[12]  Anton van den Hengel,et al.  Image-Based Recommendations on Styles and Substitutes , 2015, SIGIR.

[13]  Heiko Paulheim,et al.  Using Graph Metrics for Linked Open Data Enabled Recommender Systems , 2015, EC-Web.

[14]  Vijay V. Raghavan,et al.  Supervised Link Discovery on Large-Scale Biomedical Concept Networks , 2011, 2011 IEEE International Conference on Bioinformatics and Biomedicine.

[15]  Duncan J. Watts,et al.  Estimating the Causal Impact of Recommendation Systems from Observational Data , 2015, EC.

[16]  Joseph G. Davis,et al.  Recommendations using linked data , 2012, PIKM '12.

[17]  Dietmar Jannach,et al.  Neighborhood-restricted mining and weighted application of association rules for recommenders , 2010, ITWP@UMAP.

[18]  Bernhard Humm,et al.  A comparison of NER tools w.r.t. a domain-specific vocabulary , 2014, SEM '14.

[19]  Iván Cantador,et al.  Ontology-Based Identification of Music for Places , 2013, ENTER.

[20]  Anocha Aribarg,et al.  Category-Based Screening in Choice of Complementary Products , 2009 .

[21]  Roberto Turrin,et al.  Cross-Domain Recommender Systems , 2015, Recommender Systems Handbook.

[22]  Evgeniy Gabrilovich,et al.  A Review of Relational Machine Learning for Knowledge Graphs , 2015, Proceedings of the IEEE.

[23]  Emine Yilmaz,et al.  Towards Hierarchies of Search Tasks & Subtasks , 2015, WWW.

[24]  Jure Leskovec,et al.  Inferring Networks of Substitutable and Complementary Products , 2015, KDD.

[25]  Wei Chu,et al.  Learning to extract cross-session search tasks , 2013, WWW.

[26]  Grzegorz Chodak,et al.  Using association rules to assess purchase probability in online stores , 2016, Information Systems and e-Business Management.

[27]  Amos Azaria,et al.  Movie recommender system for profit maximization , 2013, AAAI.

[28]  Daniel Nikovski,et al.  Induction of compact decision trees for personalized recommendation , 2006, SAC.

[29]  Giuseppe Pirrò,et al.  REWOrD: Semantic Relatedness in the Web of Data , 2012, AAAI.

[30]  Ram D. Gopal,et al.  Empirical Analysis of the Impact of Recommender Systems on Sales , 2010, J. Manag. Inf. Syst..

[31]  Christopher D. Manning,et al.  Introduction to Information Retrieval Christopher D. Manning, Prabhakar Raghavan, and Hinrich Schtze (Stanford University, Yahoo! Research, and University of Stuttgart) Cambridge: Cambridge University Press, 2008, xxi+482 pp; hardbound, ISBN 978-0-521-86571-5, $60.00 , 2009, Computational Linguistics.

[32]  Zhu-Hong You,et al.  Integration of Genomic and Proteomic Data to Predict Synthetic Genetic Interactions Using Semi-supervised Learning , 2009, ICIC.

[33]  Reem Bahgat,et al.  A supervised learning approach to link prediction in Twitter , 2016, Social Network Analysis and Mining.

[34]  References , 1971 .

[35]  Junyu Niu,et al.  Substitutes or complements: another step forward in recommendations , 2009, EC '09.

[36]  Azuraliza Abu Bakar,et al.  Mining positive and Negative Association Rules from interesting frequent and infrequent itemsets , 2012, 2012 9th International Conference on Fuzzy Systems and Knowledge Discovery.