An Effective Recommendation Model using User Access Sequence and Context Entropy

An Effective Recommendation Model using User Access Sequence and Context Entropy

Xiaoyi Deng Feifei Huangfu 

College of Business Administration, Huaqiao University P.R. China, No.269, Chenghua North Rd., Quanzhou 362021

College of Foreign Languages, Huaqiao University P.R. China, No.269, Chenghua North Rd., Quanzhou 362021

Corresponding Author Email: 
londonbell@hqu.edu.cn; huangfu@hqu.edu.cn
Page: 
57-73
|
DOI: 
https://doi.org/10.18280/ama_b.600104
Received: 
15 March 2017
| |
Accepted: 
15 April 2017
| | Citation

OPEN ACCESS

Abstract: 

Collaborative filtering (CF) is one of the most successful recommendation technologies to cope with information overload problem. Conventional neighborhood-based CF models solely use user/item similarities instead of existing user preferences to form neighborhoods, which prediction accuracy excessively relies on. Besides, customers’ interests and demands may vary with contexts in different environment. As a result, the recommendations quality of conventional CF models would suffer. To address these issues, this paper developed an effective hybrid CF model by integrating user access sequence and context entropy. The user access sequence is introduced to mitigate the new user cold-start problem, and the context entropy is introduced for measuring uncertainty degree of user rating under different context and calculating user similarity for gathering the most similar users. Experiments on real-world datasets are carried out to compare our method’s accuracy with other three algorithms. The results show our method outperforms other methods and improves recommendation quality effectively.

Keywords: 

collaborative filtering, user access sequence, context entropy, cold-start, nearest neighbor selection, recommender system

1. Introduction
2. Background and Related Work
3. Hybrid Model based on User Access Sequence and Context Entropy
4. Experiments and Results
5. Conclusion
Acknowledgment

This work is supported by the National Natural Science Foundation of China (No.71401058), the Program for New Century Excellent Talents in Fujian Province University, NCETFJ (No.Z1625110), and the Project of Science and Technology Plan of Fujian Province of China (No.2017H01010065)

  References

1. X. Su, T.M. Khoshgoftaar, A survey of collaborative filtering techniques, 2009, Advances in Artificial Intelligence, vol.12, pp.1-19.

2. B. Sarwar, G. Karypis, J. Konstan, et al, Item-based collaborative filtering recommendation algorithms, 2001, International Conference on World Wide Web, Hong Kong, pp.285-295.

3. Y Koren, R Bell, Advances in collaborative filtering, 2011, Recommender Systems Handbook, Springer, pp.145-186.

4. J. Bobadilla, F. Ortega, A. Hernando, et al,  Recommender systems survey, 2013, Knowledge-Based Systems, vol.46, no.1, pp.109-132.

5. D. Ben-Shimon, L. Rokach, B. Shapira, An ensemble method for top-N recommendations from the SVD, 2016, Expert Systems with Applications, vol.64, pp.84-92.

6. J. Liu, C. Wu, Y. Xiong, et al, List-wise probabilistic matrix factorization for recommendation, 2014, Information Sciences, vol.278, pp.434-447.

7. H. Wu, K. Yue, Y. Pei, et al, Collaborative topic regression with social trust ensemble for recommendation in social media systems, 2016, Knowledge-Based Systems, vol.97, pp.111-122..

8. C. He, D. Parra, K. Verbert, Interactive recommender systems: A survey of the state of the art and future research challenges and opportunities, 2016, Expert Systems with Applications, vol.56, pp.9-27, 

9. G. Adomavicius, R. Sankaranarayanan, S. Sen, et al, Incorporating contextual information in recommender systems using a multidimensional approach, 2005, ACM Transactions on Information Systems, vol.23, no.1, 103-145.

10. D. Anand, K.K. Bharadwaj, Utilizing various sparsity measures for enhancing accuracy of collaborative recommender systems based on local and global similarities, 2011, Expert Systems with Applications, vol.38, no.5, pp.5101-5109.

11. M. Chen, C.Yang, J.Chen, et al, A method to solve cold-start problem in recommendation system based on social network sub-community and ontology decision model, 2013, The 3rd International Conference on Multimedia Technology, Guangzhou, China, pp.159-166.

12. H.J. Ahn, A new similarity measure for collaborative filtering to alleviate the new user cold-starting problem, 2008, Information Sciences, vol.178, no.1, pp.37-51.

13. M. Aharon, N. Aizenberg, E. Bortnikov, et al, OFF-set: one-pass factorization of feature sets for online recommendation in persistent cold start settings, 2013, The 7th ACM Conference on Recommender systems, Hong Kong, 2013, pp.375-378.

14. H.S. Le, Dealing with the new user cold-start problem in recommender systems: a comparative review, 2016, Information Systems, vol.58, pp.87-104.

15. F. Zhang, H. Liu, Y. Cui, Rating information entropy for cold-start recommendation, 2011, Journal of Information & Computational Science, vol.8, no.1, pp.16-22.

16. W. Wang, G. Zhang, J. Lu, Collaborative Filtering with Entropy - Driven User Similarity in Recommender Systems, 2015, International Journal of Intelligent Systems, vol.30, no.8, pp.854-870.

17. Z.D. Champiri, S.R. Shahamiri, S.S.B. Salim, A systematic review of scholar context-aware recommender systems, 2015, Expert Systems with Applications, vol.42, no.3, pp.1743-1758.

18. N. Mallat, M. Rossi, V.K. Tuunainen, et al, The impact of use context on mobile services acceptance: The case of mobile ticketing, 2009, Information & management, vol.46, no.3, pp.190-195.

19. Y. Shi, A. Karatzoglou, L. Baltrunas, et al, TFMAP: optimizing MAP for top-n context-aware recommendation, 2012, International ACM SIGIR Conference on Research and Development in Information Retrieval, Potland, USA, pp.155-164.

20. A. Said,  B. J. Jain, S. Albayrak,  Analyzing  weighting  schemes in collaborative filtering:  Cold start, post cold start and power users, 2012, The 27th Annual ACM Symposium on Applied Computing, Trento, Italy, pp.2035-2040.

21. J. Han, H.R. Schmidtke, X. Xie, et al, Adaptive content recommendation for mobile users: Ordering recommendations using a hierarchical context model with granularity, 2014, Pervasive and Mobile Computing, vol.13, pp.85-98.