Estratégia de Pós-processamento Aplicada a um Sistema de Recomendação de Artigos para a Melhora da Diversidade
Resumo
Sistemas de Recomendação buscam identificar itens relevantes para seus usuários. Um dos temas recentes na área de Sistemas de Recomendação são as questões sobre diversidade. Ao considerar aspectos para aumentar a diversidade tenta-se evitar a priorização de algumas das preferências do usuário ou ainda a recomendação de itens muito similares entre si. Este trabalho apresenta a implementação de um modelo de pós-processamento aplicado a um Sistema de Recomendação de artigos, visando a promoção da diversidade das recomendações. O sistema foi utilizado por 56 usuários de diferentes instituições, sendo a diversidade e a precisão avaliadas por métricas. O pósprocessamento aumentou a diversidade mantendo a precisão.
Palavras-chave:
sistemas de recomendação, diversidade, perfil de usuário, pós-processamento
Referências
Adomavicius, G. and Kwon, Y. (2009). Toward more diverse recommendations: Item reranking methods for recommender systems. In Workshop on Information Technologies and Systems, pages 79–84.
Antikacioglu, A. and Ravi, R. (2017). Post processing recommender systems for diversity. In Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pages 707–716.
Barraza-Urbina, A., Heitmann, B., Hayes, C., and Ramos, A. C. (2015). Xplodiv: Diversification approach for recommender systems. INSIGHT Centre for Data Analytics, National University of Ireland, Galway.
Bradley, K. and Smyth, B. (2001). Improving recommendation diversity. In Proceedings of the Twelfth Irish Conference on Artificial Intelligence and Cognitive Science, Maynooth, Ireland, volume 85, pages 141–152.
Cezar, N. L., de Borba, C., Gasparini, I., and Lichtnow, D. (2021). Applying a postprocessing strategy to consider the multiple interests of users of a paper recommender system. In Araujo, R. D., Dorça, F. A., de Araujo, R. M., Siqueira, S. W. M., and Fontão, A. L., editors, SBSI 2021: XVII Brazilian Symposium on Information Systems, Uberlândia, Brazil, June 7 10, 2021, pages 49:1–49:7. ACM.
Gormley, C. and Tong, Z. (2015). Elasticsearch: the definitive guide: a distributed realtime search and analytics engine. ”O’Reilly Media, Inc.”.
Kaminskas, M. and Bridge, D. (2016). Diversity, serendipity, novelty, and coverage: a survey and empirical analysis of beyond-accuracy objectives in recommender systems. ACM Transactions on Interactive Intelligent Systems (TiiS), 7(1):1–42.
Kaya, M. and Bridge, D. (2019). Subprofile-aware diversification of recommendations. User Modeling and User-Adapted Interaction, 29(3):661–700.
Küçüktunç, O., Saule, E., Kaya, K., and C¸ atalyürek, Ü. V. (2013). Result diversification in automatic citation recommendation. In Proceedings of the iConference workshop on Computational scientometrics: theory and applications, pages 1–4.
Kunaver, M., Dobravec, S., Pozrl, T., and Kosir, A. (2014). Increasing top-20 search results diversity through recommendation post-processing. In UMAP Workshops.
Kunaver, M. and Pozrl, T. (2017). Diversity in recommender systems – a survey. Knowledge-Based Systems, 123:154–162.
Lunardi, G. M., Machado, G. M., and de Oliveira, J. P. M. (2018). Diversificação de recomendação em cidades inteligentes: Estudo e estrutura de abordagem. Cadernos de Informática, 10(1):28–44.
Vargas, S. (2015). Novelty and diversity evaluation and enhancement in recommender systems. PhD thesis, PhD thesis, Universidad Autónoma de Madrid, Spain.
Vellino, A. (2010). A comparison between usage-based and citation-based methods for recommending scholarly research articles. Proceedings of the American Society for Information Science and Technology, 47(1):1–2.
Wang, Y., Zhang, X., Liu, Z., Dong, Z., Feng, X., Tang, R., and He, X. (2020). Personalized re-ranking for improving diversity in live recommender systems. arXiv preprint arXiv:2004.06390.
Yang, X., Guo, Y., Liu, Y., and Steck, H. (2014). A survey of collaborative filtering based social recommender systems. Computer communications, 41:1–10.
Ziegler, C.-N., McNee, S. M., Konstan, J. A., and Lausen, G. (2005). Improving recommendation lists through topic diversification. In Proceedings of the 14th international conference on World Wide Web, pages 22–32.
Antikacioglu, A. and Ravi, R. (2017). Post processing recommender systems for diversity. In Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pages 707–716.
Barraza-Urbina, A., Heitmann, B., Hayes, C., and Ramos, A. C. (2015). Xplodiv: Diversification approach for recommender systems. INSIGHT Centre for Data Analytics, National University of Ireland, Galway.
Bradley, K. and Smyth, B. (2001). Improving recommendation diversity. In Proceedings of the Twelfth Irish Conference on Artificial Intelligence and Cognitive Science, Maynooth, Ireland, volume 85, pages 141–152.
Cezar, N. L., de Borba, C., Gasparini, I., and Lichtnow, D. (2021). Applying a postprocessing strategy to consider the multiple interests of users of a paper recommender system. In Araujo, R. D., Dorça, F. A., de Araujo, R. M., Siqueira, S. W. M., and Fontão, A. L., editors, SBSI 2021: XVII Brazilian Symposium on Information Systems, Uberlândia, Brazil, June 7 10, 2021, pages 49:1–49:7. ACM.
Gormley, C. and Tong, Z. (2015). Elasticsearch: the definitive guide: a distributed realtime search and analytics engine. ”O’Reilly Media, Inc.”.
Kaminskas, M. and Bridge, D. (2016). Diversity, serendipity, novelty, and coverage: a survey and empirical analysis of beyond-accuracy objectives in recommender systems. ACM Transactions on Interactive Intelligent Systems (TiiS), 7(1):1–42.
Kaya, M. and Bridge, D. (2019). Subprofile-aware diversification of recommendations. User Modeling and User-Adapted Interaction, 29(3):661–700.
Küçüktunç, O., Saule, E., Kaya, K., and C¸ atalyürek, Ü. V. (2013). Result diversification in automatic citation recommendation. In Proceedings of the iConference workshop on Computational scientometrics: theory and applications, pages 1–4.
Kunaver, M., Dobravec, S., Pozrl, T., and Kosir, A. (2014). Increasing top-20 search results diversity through recommendation post-processing. In UMAP Workshops.
Kunaver, M. and Pozrl, T. (2017). Diversity in recommender systems – a survey. Knowledge-Based Systems, 123:154–162.
Lunardi, G. M., Machado, G. M., and de Oliveira, J. P. M. (2018). Diversificação de recomendação em cidades inteligentes: Estudo e estrutura de abordagem. Cadernos de Informática, 10(1):28–44.
Vargas, S. (2015). Novelty and diversity evaluation and enhancement in recommender systems. PhD thesis, PhD thesis, Universidad Autónoma de Madrid, Spain.
Vellino, A. (2010). A comparison between usage-based and citation-based methods for recommending scholarly research articles. Proceedings of the American Society for Information Science and Technology, 47(1):1–2.
Wang, Y., Zhang, X., Liu, Z., Dong, Z., Feng, X., Tang, R., and He, X. (2020). Personalized re-ranking for improving diversity in live recommender systems. arXiv preprint arXiv:2004.06390.
Yang, X., Guo, Y., Liu, Y., and Steck, H. (2014). A survey of collaborative filtering based social recommender systems. Computer communications, 41:1–10.
Ziegler, C.-N., McNee, S. M., Konstan, J. A., and Lausen, G. (2005). Improving recommendation lists through topic diversification. In Proceedings of the 14th international conference on World Wide Web, pages 22–32.
Publicado
31/07/2022
Como Citar
SOUZA, Ediana da Silva de; LICHTNOW, Daniel; GASPARINI, Isabela.
Estratégia de Pós-processamento Aplicada a um Sistema de Recomendação de Artigos para a Melhora da Diversidade. In: BRAZILIAN WORKSHOP ON SOCIAL NETWORK ANALYSIS AND MINING (BRASNAM), 11. , 2022, Niterói.
Anais [...].
Porto Alegre: Sociedade Brasileira de Computação,
2022
.
p. 216-221.
ISSN 2595-6094.
DOI: https://doi.org/10.5753/brasnam.2022.222805.
