Using principles of locality and connectivity of the context in recommender systems
DOI:
https://doi.org/10.20998/2079-0023.2018.22.03Keywords:
recommender systems, ranking of results, decision making context, reference systems, locality, connectivityAbstract
The problem of the relevance of input data in advisory systems is investigated. This problem arises due to insufficient differentiation of data on goods relative to consumers, which does not allow to fully individualize their preferences in the advisory system. To solve this problem, it is suggested to take into account the local contexts of consumers, reflecting the conditions for the acceptance of the choice by these consumers. Using the context allows you to set contextual constraints on possible variants of an ordered list of recommendations and thereby improve the quality of the recommendation system. In order to provide context-oriented recommendations, it is proposed to consistently generalize and filter out the local contexts of consumers using the principles of locality and connectivity. The peculiarity of using these principles is that the static and dynamic aspects of the context are combined. The first aspect is characterized by a set of properties of objects that are of interest to the consumer. The second aspect is given in the form of patterns of events reflecting the consumer’s behavior with respect to these objects. The proposed relationship between the aspects is that each event corresponds to a pair of successive sets of object properties that differ in one property value. A two-phase approach to the formation of a decision-making context for a recommendation system is proposed, which provides for the consistent integration of the static and dynamic components of the context. Integration uses an equivalence, similarity and compatibility relationship. When the first phase is implemented, item-based is formed, and the second is a user-based context description. Then these descriptions are combined and filtered in accordance with the characteristics of the new consumer to whom the recommendations are issued. The practical significance of the proposed approach is that it allows you to delete irrelevant input data taking into account the context of the decision-making by the consumer and, on this basis, improve the accuracy of the recommendations.References
Aggarwal C. C. Recommender Systems: The Textbook. Springer, New York, 2017. 498 p.
Abowd G., Atkeson C., Hong J., Long S., Kooper R., Pinkerton M. Cyberguide: A mobile context-aware tour guide. Wireless Networks. 1997, no. 3(5), pp. 421–433.
Herlocker J.L., Konstan J.A., Terveen L.G., Riedl J.T. Evaluating collaborative filtering recommender systems. ACM Transactions on Information Systems (TOIS). 2004, vol. 22, no. 1, pp. 5–53.
Abowd G., Dey A., Brown P., Davies N., Smith M., Steggles P. Towards a better understanding of context and context-awareness. Handheld and Ubiquitous Computing. 1999, pp. 304–307.
Adomavicius G., Sankaranarayanan R., Sen S., Tuzhilin A. Incorporating contextual information in recommender systems using a multidimensional approach. ACM Transactions on Information Systems. 2005, no. 23(1), pp. 103–145.
Linden G., Smith B., York J. Amazon.com recommendations: Itemto-item collaborative filtering, Internet Computing, IEEE. 2003, vol. 7, no.1, pp. 76–80.
Adomavicius G., Tuzhilin A. Context-aware recommender systems. Recommender Systems handbook. Springer, NY, 2011, pp. 217– 253.
Adomavicius G. Tuzhilin A. Incorporating context into recommender systems using multidimensional rating estimation methods. International Workshop on Web Personalization, Recommender Systems and Intelligent User Interfaces (WPRSIUI). 2005, pp. 3–13.
Shaw Gavin, Xu Yue. Using Association Rules to Solvethe ColdStart Problem in Recommender Systems. [QUT Digital Repository]. Available at: http://eprints.qut.edu.au/40176 (accessed 24.05.2018).
Sobhanam Hridya, Mariappan A.K. Addressing cold start problem in recommender systems using association rules and clustering technique. International Conference on Computer Communication and Informatics (ICCCI- 2013). Coimbatore, India, 2013, pp. 402– 411.
Adomavicius G., Tuzhilin A. Towards the Next Generation of Recommender Systems. A Survey of the State-of-the-Art and Possible Extensions. IEEE Transactions on Knowledge and Data Engineering, 2005, no. 17, pp. 634–749.
Baltrunas L. Ludwig B. Peer S., Ricci F. Context- Aware Places of Interest Recommendations for Mobile Users. Proceedings of the 14th International Conference on Human-Computer Interaction. Berlin, Springer, 2011, pp. 531–540.
Agarwal D., Chen B. C., Long B. Localized factor models for multicontext recommendation. ACM KDD Conference. 2011, pp. 609– 617.
Chalaya O.V. Pryntsyp ta metod evoliutsiinoi pobudovy bazy znan’ na osnovi analizu logiv IS protsesnogo upravlinnia [Development of knowledge base after results of analysis of the logs of the process management information system]. Naukovo-tekhnichnyi zhurnal «Bionika intelektu» [ Scientific and Technical Journal "Bionics of Intellect"]. Kharkiv, NURE Publ., 2017, no. 1(88), pp. 80–84.
Downloads
How to Cite
Issue
Section
License
Copyright (c) 2018 Bulletin of National Technical University "KhPI". Series: System Analysis, Control and Information TechnologiesAuthors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).