融合Jensen-Shannon散度的推荐算法
|
摘要
为充分利用所有评分,缓解数据稀疏性问题,将概率统计领域的Jensen-Shannon(JS)散度引入相似性度量中,提出了一种新的项目相似性度量算法。该算法将项目的评分信息转化为评分值密度,并依据评分值的密度分布来计算项目相似性。同时,引入评分数量因子,进一步提升了基于JS的相似性度量方法的性能。最后,以基于JS的相似性度量方法为基础,设计了相应的协同过滤算法。在MovieLens数据集上的实验结果表明,所提算法在预测误差和推荐准确性方面均有良好的表现。因此,该算法在推荐系统中具有很好的应用潜力。
To fully utilize all the ratings and weaken the problem of data sparsity,the Jensen-Shannon divergence in statistics field was used to design a new similarity measure for items.In this similarity measure,the ratings for items are converted to the density of rating values.Then,the item similarity is calculated according to the density of rating values.Meanwhile,the factor for the number of ratings is also considered to further enhance the performance of the proposed similarity measure based on JS divergence.Finally,a collaborative filtering recommendation algorithm is presented according to the JS-divergence-based item similarity.The test results on MovieLens dataset show that the proposed algorithm has good performance in prediction error and recommendation precision.Therefore,it has high potential to be applied in recommendation system.
To fully utilize all the ratings and weaken the problem of data sparsity,the Jensen-Shannon divergence in statistics field was used to design a new similarity measure for items.In this similarity measure,the ratings for items are converted to the density of rating values.Then,the item similarity is calculated according to the density of rating values.Meanwhile,the factor for the number of ratings is also considered to further enhance the performance of the proposed similarity measure based on JS divergence.Finally,a collaborative filtering recommendation algorithm is presented according to the JS-divergence-based item similarity.The test results on MovieLens dataset show that the proposed algorithm has good performance in prediction error and recommendation precision.Therefore,it has high potential to be applied in recommendation system.
引文
[1] CHOU A Y.The analysis of online social networking:How technology is changing e-commerce purchasing decision[J].International Journal of Information Systems&Change Management,2010,4(4):353-365.
[2] YANG C C.Correlation coefficient evaluation for the fuzzy interval data[J].Journal of Business Research,2016,69(6):2138-2144.
[3] GUAN H,GUAN S,ZHAO A.Forecasting Model Based on Neutrosophic Logical Relationship and Jaccard Similarity[J].Symmetry,2017,9(9):191.
[4] TAKACS G,PILASZY I,NEMETH B,et al.Scalable Collaborative Filtering Approaches for Large Recommender System[J].Journal of Machine Learning Research,2009,10:623-656.
[5] KIM H N,JI A T,HA I,et al.Collaborative Filtering Based on Collaborative Tagging for Enhancing the Quality of Recommendation[J].Electronic Commerce Research and Applications,2010,9(1):73-83.
[6] KHROUF H.Hybrid event recommendation using linked data and user diversity[C]∥ACM Conference on Recommender Systems.ACM,2013:185-192.
[7] MEYMANDPOUR R,DAVIS J G.Recommendations using linked data[C]∥Proceedings of the 5th Ph.d.Workshop on Information and Knowledge.ACM,2012:75-82.
[8] OSTUNI V C,NOIA T D,SCIASCIO E D,et al.Top-N recommendations from implicit feedback leveraging linked open data[C]∥ACM Conference on Recommender Systems.ACM,2013:85-92.
[9] BARJASTEH I,FORSATI R,MASROUR F,et al.Cold-Start Item and User Recommendation with Decoupled Completion and Transduction[C]∥ACM Conference on Recommender Systems.ACM,2015:91-98.
[10]BINESH N,REZGHI M.A new similarity measure for extraction information from social networks and improve the community detection and recommendation results[C]∥Information and Knowledge Technology.IEEE,2015:146-151.
[11]WANG X M,ZHANG X M,WU Y T,et al.A Collaborative Recommendation Algorithm Based on Heuristic Clustering Model and Category Similarity[J].Acta Electronica Sinica,2016,44(7):1708-1713.(in Chinese)王兴茂,张兴明,吴毅涛,等.基于启发式聚类模型和类别相似度的协同过滤推荐算法[J].电子学报,2016,44(7):1708-1713.
[12]WANG Y,DENG J Z,DENG Y H,et al.A Collaborative Filtering Recommendation Algorithm Based on Item Probability Distribution[J].New Technology of Library and Information Service,2016,32(6):73-79.(in Chinese)王永,邓江洲,邓永恒,等.基于项目概率分布的协同过滤推荐算法[J].现代图书情报技术,2016,32(6):73-79.
[13]PATRA B K,LAUNONEN R,OLLIKAINEN V,et al.A new similarity measure using Bhattacharyya coefficient for collaborative filtering in sparse data[J].Knowledge-Based Systems,2015,82(3):163-177.
[14]MANNING C D.Foundations of statistical natural language processing[M].Massachusetts:MIT Press,1999.
[15]MAJTEY A P,LAMBERTI P W,PRATO D P.Jensen-Shannon divergence as a measure of distinguishability between mixed quantum states[J].Physical Review A,2005,72(5):762-776.
[16]WILLMOTT C J,MATSUURA K.Advantages of the Mean Absolute Error(MAE)over the Root Mean Square Error(RMSE)in Assessing Average Model Performance[J].Climate Research,2005,30(1):79-82.
1)http://www.grouplens.org
[2] YANG C C.Correlation coefficient evaluation for the fuzzy interval data[J].Journal of Business Research,2016,69(6):2138-2144.
[3] GUAN H,GUAN S,ZHAO A.Forecasting Model Based on Neutrosophic Logical Relationship and Jaccard Similarity[J].Symmetry,2017,9(9):191.
[4] TAKACS G,PILASZY I,NEMETH B,et al.Scalable Collaborative Filtering Approaches for Large Recommender System[J].Journal of Machine Learning Research,2009,10:623-656.
[5] KIM H N,JI A T,HA I,et al.Collaborative Filtering Based on Collaborative Tagging for Enhancing the Quality of Recommendation[J].Electronic Commerce Research and Applications,2010,9(1):73-83.
[6] KHROUF H.Hybrid event recommendation using linked data and user diversity[C]∥ACM Conference on Recommender Systems.ACM,2013:185-192.
[7] MEYMANDPOUR R,DAVIS J G.Recommendations using linked data[C]∥Proceedings of the 5th Ph.d.Workshop on Information and Knowledge.ACM,2012:75-82.
[8] OSTUNI V C,NOIA T D,SCIASCIO E D,et al.Top-N recommendations from implicit feedback leveraging linked open data[C]∥ACM Conference on Recommender Systems.ACM,2013:85-92.
[9] BARJASTEH I,FORSATI R,MASROUR F,et al.Cold-Start Item and User Recommendation with Decoupled Completion and Transduction[C]∥ACM Conference on Recommender Systems.ACM,2015:91-98.
[10]BINESH N,REZGHI M.A new similarity measure for extraction information from social networks and improve the community detection and recommendation results[C]∥Information and Knowledge Technology.IEEE,2015:146-151.
[11]WANG X M,ZHANG X M,WU Y T,et al.A Collaborative Recommendation Algorithm Based on Heuristic Clustering Model and Category Similarity[J].Acta Electronica Sinica,2016,44(7):1708-1713.(in Chinese)王兴茂,张兴明,吴毅涛,等.基于启发式聚类模型和类别相似度的协同过滤推荐算法[J].电子学报,2016,44(7):1708-1713.
[12]WANG Y,DENG J Z,DENG Y H,et al.A Collaborative Filtering Recommendation Algorithm Based on Item Probability Distribution[J].New Technology of Library and Information Service,2016,32(6):73-79.(in Chinese)王永,邓江洲,邓永恒,等.基于项目概率分布的协同过滤推荐算法[J].现代图书情报技术,2016,32(6):73-79.
[13]PATRA B K,LAUNONEN R,OLLIKAINEN V,et al.A new similarity measure using Bhattacharyya coefficient for collaborative filtering in sparse data[J].Knowledge-Based Systems,2015,82(3):163-177.
[14]MANNING C D.Foundations of statistical natural language processing[M].Massachusetts:MIT Press,1999.
[15]MAJTEY A P,LAMBERTI P W,PRATO D P.Jensen-Shannon divergence as a measure of distinguishability between mixed quantum states[J].Physical Review A,2005,72(5):762-776.
[16]WILLMOTT C J,MATSUURA K.Advantages of the Mean Absolute Error(MAE)over the Root Mean Square Error(RMSE)in Assessing Average Model Performance[J].Climate Research,2005,30(1):79-82.
1)http://www.grouplens.org