An Enhanced Framework for Efficient Point-of-Interest Recommendations in Recommender Systems
محورهای موضوعی : پردازش چند رسانه ای، سیستمهای ارتباطی، سیستمهای هوشمندَAli Asghar Salehi Soleymanabadi 1 , Keyhan Khamforoosh 2 * , Vafa Maihami 3
1 - PhD Student, Department of Computer Engineering, Sa.C., Islamic Azad University, Sanandaj, Iran
2 - Assistant Professor, Department of Computer Engineering, Sa.C., Islamic Azad University, Sanandaj, Iran
3 - Assistant Professor, Department of Computer Engineering, Sa.C., Islamic Azad University, Sanandaj, Iran
کلید واژه: سیستم های توصیه گر, نقاط علاقه, فریم ورک اصلاح شده , نوتروسفیک,
چکیده مقاله :
امروزه سیستمهای توصیهگر یکی از مهمترین موضوعات مورد تحقیق است.سیستمهای توصیهگر مبتنی بر مکان با استفاده از دادههای مکانی و تحلیل رفتار کاربران، نقش مهمی در بهبود تجربه کاربران در حوزههایی مانند پیشنهاد رستورانها، مکانهای تفریحی و خدمات شهری دارند. با رشد فناوریهای هوش مصنوعی و افزایش حجم دادههای مکانی، توسعه روشهای بهینهسازی این سیستمها ضروری شده است. این پژوهش یک فریمورک نوین برای بهبود دقت و کیفیت پیشنهادات ارائه میکند که مبتنی بر روش های یادگیری ماشین در اجزای فربم ورک و مدلسازی تصمیمگیری چندمعیاره نوتروسفیک است.
مدل پیشنهادی شامل سه مرحله اصلی است: پردازش دادههای مکانی، بهینهسازی مدل توصیهگر و رتبهبندی پیشنهادات نهایی. برای تحلیل دقیقتر مکانها، ویژگی های کلیدی از جمله مجاورت جغرافیایی، الگوهای رفتاری کاربران و دادههای شبکههای اجتماعی شناسایی شدهاند. الگوریتمهای خوشهبندی مانند رقابت استعماری و منطق فازی C-MEANS کلاسترهای نزدیک به هم را از نظر جغرافیایی پیدا می کند . در کنار آن، روش تصمیمگیری چندمعیاره ویکور نوتروسفیک برای رتبهبندی نهایی در هر کلاستر استفاده شده است.
نتایج ارزیابی فریمورک پیشنهادی با مجموعه دادههای Yelp، Foursquare و Flickr نشان میدهد که این مدل نسبت به روشهای سنتی دقت بالاتری دارد. مقایسه این روش با روشهای پایه مانند Popularity Rank (PR)، Classic Rank (CLR) و Frequent Rank (FR) نشاندهنده بهبود عملکرد در معیارهای دقت، میانگین قدرمطلق خطا و نرخ پذیرش پیشنهادات توسط کاربران است. همچنین، مقایسه با پژوهشهای دیگر نشان میدهد که استفاده از دادههای جامعتر مانند زمان بازدید، شرایط آب و هوا و تحلیل دقیقتر دادهها موجب بهبود کیفیت توصیهها شده است.
در نهایت، مقایسه روش پیشنهادی با الگوریتمهای CRPNS ،MMPOI که هر یک فریم ورکی را پیشنهاد داده اند نشان میدهد که بهبود های انحام شده، در معیارهای Precision، Recall و NDCG عملکرد بهتری دارد. این یافتهها نشان میدهد که بهرهگیری از مدل نوتروسفیک در تصمیمگیری چندمعیاره میتواند بهطور مؤثری دقت و کیفیت پیشنهادات را بهبود بخشد.
Location-based recommender systems significantly enhance user experience in areas such as restaurant recommendations, recreational spaces, and urban services by utilizing spatial data and analyzing user behavior. This research proposes a novel framework to improve the accuracy and quality of recommendations, leveraging machine learning methods and neutrosophic multi-criteria decision modeling. The proposed model consists of three main stages: spatial data processing, recommender system optimization, and final recommendation ranking. Key features, including geographical proximity, user behavioral patterns, and social network data, have been identified to analyze locations more accurately. Clustering algorithms, such as the Imperialist Competitive Algorithm (ICA) and Fuzzy C-Means, are employed to identify geographically close clusters. Additionally, the neutrosophic VIKOR multi-criteria decision-making method is utilized for final ranking within each cluster. Evaluation results using datasets from Yelp, Foursquare, and Flickr indicate that the proposed model achieves higher accuracy compared to traditional methods. A comparison with baseline approaches, including Popularity Rank (PR), Classic Rank (CLR), and Frequent Rank (FR), demonstrates improved performance in terms of accuracy, mean absolute error, and user acceptance rate of recommendations. Furthermore, comparisons with other studies reveal that incorporating more comprehensive data, such as visit time and weather conditions, along with more detailed data analysis, enhances the quality of recommendations. Finally, a comparison of the proposed method with the CPRNS and MMPOI algorithms, both of which introduced a framework, shows that the implemented improvements achieve better performance in terms of Precision, Recall, and NDCG2 metrics.
[1] Ye, M., Yin, P., Lee, W.C., Lee, D.L.: Exploiting geographical influence for collaborative point-of-interest recommendation, 325–334 (2011)
[2] Zhu, J.H., Ming, Q.: point-of-interest recommendation by incorporating trust-distrust relationship in lbsn. Journal on Communications 39(7), 157–165 (2018)
[3] Rashidi, R., Khamforoosh, K., & Sheikhahmadi, A. (2020). An analytic approach to separate users by introducing new combinations of initial centers of clustering. Physica A: statistical mechanics and its applications, 551, 124185.
[4] Si, Y., Zhang, F., Liu, W.: An adaptive point-of-interest recommendation method for location-based social networks based on user activity and spatial features. Knowledge-Based Systems 163, 267–282 (2019).doi:10.1016/j.knosys.2018.08.031
[5] Liu, S., Wang, L.: A self-adaptive point-of-interest recommendation algorithm based on a multi-order Markov model. Future Generation Computer Systems 89, 506–514 (2018)
[6] Xing, S., Liu, F., Wang, Q., Zhao, X., Li, T.: Content-aware point-of-interest recommendation based on convolutional neural network. Applied Intelligence 49(3), 858–871 (2019)
[7] Smarandache, F.: A unifying field in logics: Neutrosophic logic. Multiple-Valued Logic 8(3), 489–503 (2002).
[8] L. Liu, N. Mehandjiev, and D.-L. Xu, ``Multi-criteria service recommendation based on user criteria preferences,'' in Proc. 5th ACM Conf. Recommender Syst., 2011, pp. 77_84.
[9] F. Hdioud, B. Frikh, and B. Ouhbi, ``Multi-criteria recommender systems based on multi-attribute decision making,'' in Proc. Int. Conf. Inf. Integr. Web-Based Appl. Services, 2013, p. 203.
[10] Han, J., Yamana, H.: Geographic diversification of recommended POI's in frequently visited areas. In: ACM Transactions on Information Systems, vol. 38, pp. 1–39 (2019)
[11] Cheng, C., Yang, H., King, I., Lyu, M.: Fused matrix factorization with geographical and social influence in location-based social networks. Proceedings of the National Conference on Artificial Intelligence, 17–23 (2012)
[12] Capdevila, J., Arias, M., Arratia, A.: Geosrs : A hybrid social recommender system for geolocated data. Information Systems 57, 111–128 (2016)
[13] Ren, X.Y., Song, M.N., Song, J.D.: point-of-interest recommendation based on the user check-in behavior. Chinese Journal of Computers 40(1)
[14] Liu, Y., Seah, H.S.: Points of interest recommendation from gps trajectories. International Journal of Geographical Information Systems 29(6), 953–979 (2015)
[15] Yin, H., Cui, B., Zhou, X., Wang, W., Zi, H., Sadiq, S.: Joint modeling of user check-in behaviors for real-time Point-of-interest recommendation. In: Acm International on Conference on Information and Knowledge Management, vol. 35, pp. 1631–1640 (2015)
[16] Davtalab, M., Alesheikh, A.A.: A POI recommendation approach integrating social spatio-temporal information into probabilistic matrix factorization. Knowledge and Information Systems 63(1), 65–85 (2021)
[17] Cheng, C., Yang, H., King, I., Lyu, M.: Fused matrix factorization with geographical and social influence in location-based social networks. Proceedings of the National Conference on Artificial Intelligence, 17–23 (2012)
[18] Zhang, J.: Geosoca : Exploiting geographical,social and categorical correlations for point-of-interest recommendations categories and subject descriptors. ACM Special Interest Group on Information Retrieval, 443–452 (2015)
[19] Liu, B., Xiong, H., Papadimitriou, S., Fu, Y., Yao, Z.: A general geographical probabilistic factor model for point of interest recommendation. IEEE Transactions on Knowledge & Data Engineering 27(5), 1167–1179 (2015)
[20] Gholami, M., Sheikhahmadi, A., Khamforoosh, K., Jalili, M., & Veisi, F. (2024). A new approach to finding overlapping community structure in signed networks based on Neutrosophic theory. Journal of Complex Networks, 12(1), cnad051.
[21] Zhang, H.Y., Pu, J., Wang, J.Q., Chen, X.H.: A novel decision support model for satisfactory restaurants utilizing social information: A case study of tripadvisor.com. Tourism Management 59, 281–297 (2017)
[22] Hua, M., Hu, Z., Li, K., Zhang, H.: Toward trustworthy cloud service selection: A time-aware approach using interval neutrosophic set. Journal of Parallel and Distributed Computing 96(C), 75–94 (2016)
[23] Sabaghian K, Khamforoosh K, Ghaderzadeh A (2021) Presentation of a new method based on modern multivariate approaches for big data replication in distributed environments. PLoS ONE 16(7): e0254210. https://doi.org/10.1371/journal.pone.0254210
[24] Yang Xu, Gao Cong, Lei Zhu, and Lizhen Cui. 2024. MMPOI: A Multi-Modal Content-Aware Framework for POI Recommendations. In Proceedings of the ACM Web Conference 2024 (WWW '24). Association for Computing Machinery, New York, NY, USA, 3454–3463. https://doi.org/10.1145/3589334.3645449
[25] T. Wang and C. Wang, "Embracing LLMs for Point-of-Interest Recommendations" in IEEE Intelligent Systems, vol. 39, no. 01, pp. 56-59, Jan.-Feb. 2024, doi: 10.1109/MIS.2023.3343489.
[26] M. O. Ayemowa, R. Ibrahim and M. M. Khan, "Analysis of Recommender System Using Generative Artificial Intelligence: A Systematic Literature Review," in IEEE Access, vol. 12, pp. 87742-87766, 2024, doi: 10.1109/ACCESS.2024.3416962.
[27] Zheng M, Li Y, He Z, et al. Context-aware POI Recommendation using Neutrosophic Set for Mobile Edge Computing. Research Square; 2021. DOI: 10.21203/rs.3.rs-1197954/v1.
[28] Salehi Solaiman Abadi AA, Khamforoosh K, Maihami V (2025) Point-of-interest recommender model using geo-tagged photos in accordance with imperialist Fuzzy C-means clustering. PLoS ONE 20(3): e0317131. https://doi.org/10.1371/journal.pone.0317131
[29] Atanassov,K.T. 1999. Intuitionistic Fuzzy Sets, Pysica-Verlag A Springer-Verlag Company. New York.
[30] Ban, A. 2008. “Trapezoidal approximations of intuitionistic fuzzy numbers expressed by value, ambiguity, width and weighted expected value.” Twelfth Int. Conf. on IFSs, Sofia, NIFS Vol. 14 (1):38–47.
[31] Chen,T.Y., H.P. Wang, Y.Y. Lu. 2011. “A multicriteria group decision-making approach based on interval-valued intuitionistic fuzzy sets: a comparative perspective.” Exp. Syst. Appl. 38 (6): 7647-7658.