A User and Item-Based, Wide and Deep Learning Based Travel Recommendation System

Öz

With the development of technology, the increasing amount of digital information and the widespread use of the internet, and the increase in the number of websites where trade transactions such as products, services and subscriptions are carried out on the internet, along with it, personalized and accurate; The importance of recommending products, services and subscriptions has also increased. Product-based, user-based, and hybrid traditional approaches, which are widely used in recommendations to customers, are used in most studies. Overcoming the problems of traditional approaches such as working with large and sparse data, inability to find complex relationships between user and product, and cold start has been possible with the use of deep and wide learning systems. Within the scope of this study, a comprehensive view of deep and wide neural networks and their applications in travel recommendation systems is presented. After giving the basic information about deep learning, Google's Wide and Deep Algorithm and Facebook's Deep Learning Recommendation Model (DLRM) algorithm, which are the most popular recommendation algorithms, are included. Within the scope of this study, a new travel recommendation system was created in which numerical data is fed to the model by applying the embedding process of categorical user and product features with a broad and deep learning approach. The proposed method was applied on the data set of a real travel agency company. As a result of this study, 82.37% accuracy rate was achieved in the top five recommendations given to the users.

Anahtar Kelimeler

• recommendation systems
• deep learning
• deep and wide recommendation systems
• content-based
• collaborative
• hybrid
• travel recommendation systems.

Kullanıcı ve Öğe Bazlı, Geniş ve Derin Öğrenme Tabanlı Seyahat Öneri Sistemi

Öz

Teknolojinin gelişmesi ile birlikte artan dijital bilgi miktarı ve internetin yaygınlaşması ile internet üzerinden ürün, hizmet, abonelik gibi ticaret işlemlerinin gerçekleştiği web sitelerinin sayısının da artması, beraberinde, müşterilere kişiselleştirilmiş ve doğru; ürün, hizmet ve abonelikleri sunmanın (önermenin) de önemini artmıştır. Müşterilere önerilerde yaygın olarak kullanılan ürün bazlı, kullanıcı tabanlı ve bu ikisinin birlikte kullanıldığı hibrit geleneksel yaklaşımlar çoğu çalışmada kullanılmaktadır. Geleneksel yaklaşımların, büyük ve seyrek veriler ile çalışma, kullanıcı ve ürün arasındaki karışık ilişkileri bulamama ve soğuk başlangıç (cold start) gibi problemlerinin üstesinden gelmek, derin ve geniş öğrenme sistemlerinin kullanımı ile mümkün olmuştur. Bu çalışma kapsamında, derin ve geniş sinir ağlarına ve bunların seyahat öneri sistemlerindeki uygulamalarına kapsamlı bir bakış açısı sunulmuştur. Derin öğrenme ile ilgili temel bilgilere yer verildikten sonra, en popüler öneri algoritmaları olan Google'ın Geniş ve Derin Algoritması ve Facebook'un Deep Learning Recommendation Model (DLRM) algoritmasına yer verilmiştir. Bu çalışma kapsamında, geniş ve derin öğrenme yaklaşımı ile kullanıcı ve ürün özelliklerinin kategorik olanlarının gömme işlemi uygulanarak, nümerik veriler ile modele beslendiği yeni bir seyahat öneri sistemi oluşturulmuştur. Önerilen yöntem gerçek bir seyahat acentesi şirketinin veri seti üzerinde uygulanmıştır. Bu çalışma sonucunda kullanıcılara verilen en iyi beş öneride, %82.37 doğruluk oranı yakalanmıştır.

Anahtar Kelimeler

• öneri sistemleri
• derin öğrenme
• derin ve geniş öneri sistemleri
• içerik tabanlı
• işbirlikçi
• hibrit
• seyahat öneri sistemleri

Teşekkür

Bu çalışmada veri sağlayıcısı olan BiletBank'a desteklerinden dolayı teşekkür ederiz.

Kaynakça

1. Cheng, H.-T., Koc, L., Harmsen, J., Shaked, T., Chandra, T., Aradhye, H., Anderson, G., Corrado, G., Chai, W., Ispir, M., Anil, R., Haque, Z., Hong, L., Jain, V., Liu, X., & Shah, H. (2016). Wide & Deep Learning for Recommender Systems. Içinde Proceedings of the 1st Workshop on Deep Learning for Recommender Systems. DLRS 2016: Workshop on Deep Learning for Recommender Systems. ACM. https://doi.org/10.1145/2988450.2988454
2. Naumov, M., Mudigere, D., Shi, H.-J. M., Huang, J., Sundaraman, N., Park, J., Wang, X., Gupta, U., Wu, C.-J., Azzolini, A. G., Dzhulgakov, D., Mallevich, A., Cherniavskii, I., Lu, Y., Krishnamoorthi, R., Yu, A., Kondratenko, V., Pereira, S., Chen, X., … Smelyanskiy, M. (2019). Deep Learning Recommendation Model for Personalization and Recommendation Systems (Versiyon 1). arXiv. https://doi.org/10.48550/ARXIV.1906.00091
3. Fan, W., Zhao, X., Chen, X., Su, J., Gao, J., Wang, L., Liu, Q., Wang, Y., Xu, H., Chen, L., & Li, Q. (2022). A Comprehensive Survey on Trustworthy Recommender Systems (Versiyon 1). arXiv. https://doi.org/10.48550/ARXIV.2209.10117
4. Karimova, F. (2016). A Survey of e-Commerce Recommender Systems. Içinde European Scientific Journal, ESJ (C. 12, Issue 34, s. 75). European Scientific Institute, ESI. https://doi.org/10.19044/esj.2016.v12n34p75
5. Belluf, T., Xavier, L., & Giglio, R. (2012). Case study on the business value impact of personalized recommendations on a large online retailer. Içinde Proceedings of the sixth ACM conference on Recommender systems. RecSys ’12: Sixth ACM Conference on Recommender Systems. ACM. https://doi.org/10.1145/2365952.2366014
6. Amatriain, X., & Basilico, J. (2016). Past, Present, and Future of Recommender Systems. Içinde Proceedings of the 10th ACM Conference on Recommender Systems. RecSys ’16: Tenth ACM Conference on Recommender Systems. ACM. https://doi.org/10.1145/2959100.2959144
7. Koren, Y. (2008). Factorization meets the neighborhood. Içinde Proceedings of the 14th ACM SIGKDD international conference on Knowledge discovery and data mining. KDD08: The 14th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. ACM. https://doi.org/10.1145/1401890.1401944
8. Rendle, S. (2010). Factorization Machines. Içinde 2010 IEEE International Conference on Data Mining. 2010 IEEE 10th International Conference on Data Mining (ICDM). IEEE. https://doi.org/10.1109/icdm.2010.127

9. Paparrizos, I., Cambazoglu, B. B., & Gionis, A. (2011). Machine learned job recommendation. Içinde Proceedings of the fifth ACM conference on Recommender systems. RecSys ’11: Fifth ACM Conference on Recommender Systems. ACM. https://doi.org/10.1145/2043932.2043994
10. T. Al-Otaibi, S. (2012). A survey of job recommender systems. Içinde International Journal of the Physical Sciences (C. 7, Issue 29). Academic Journals. https://doi.org/10.5897/ijps12.482
11. Albadvi, A., & Shahbazi, M. (2009). A hybrid recommendation technique based on product category attributes. Içinde Expert Systems with Applications (C. 36, Issue 9, ss. 11480-11488). Elsevier BV. https://doi.org/10.1016/j.eswa.2009.03.046
12. Balabanović, M., & Shoham, Y. (1997). Fab: content-based, collaborative recommendation. Içinde Communications of the ACM (C. 40, Issue 3, ss. 66-72). Association for Computing Machinery (ACM). https://doi.org/10.1145/245108.245124
13. Pizzato, L., Rej, T., Chung, T., Koprinska, I., & Kay, J. (2010). RECON a reciprocal recommender for online dating. Içinde Proceedings of the fourth ACM conference on Recommender systems. RecSys ’10: Fourth ACM Conference on Recommender Systems. ACM. https://doi.org/10.1145/1864708.186474
14. Bengio, Y., Lecun, Y., & Hinton, G. (2021). Deep learning for AI. Içinde Communications of the ACM (C. 64, Issue 7, ss. 58-65). Association for Computing Machinery (ACM). https://doi.org/10.1145/3448250
15. Peng, Y., Zhu, W., Zhao, Y., Xu, C., Huang, Q., Lu, H., Zheng, Q., Huang, T., & Gao, W. (2017). Cross-media analysis and reasoning: advances and directions. Içinde Frontiers of Information Technology & Electronic Engineering (C. 18, Issue 1, ss. 44-57). Zhejiang University Press. https://doi.org/10.1631/fitee.1601787
16. Wu, C.-Y., Ahmed, A., Beutel, A., Smola, A. J., & Jing, H. (2017). Recurrent Recommender Networks. Içinde Proceedings of the Tenth ACM International Conference on Web Search and Data Mining. WSDM 2017: Tenth ACM International Conference on Web Search and Data Mining. ACM. https://doi.org/10.1145/3018661.3018689
17. Elkahky, A. M., Song, Y., & He, X. (2015). A Multi-View Deep Learning Approach for Cross Domain User Modeling in Recommendation Systems. Içinde Proceedings of the 24th International Conference on World Wide Web. WWW ’15: 24th International World Wide Web Conference. International World Wide Web Conferences Steering Committee. https://doi.org/10.1145/2736277.2741667
18. Song, Y., Elkahky, A. M., & He, X. (2016). Multi-Rate Deep Learning for Temporal Recommendation. Içinde Proceedings of the 39th International ACM SIGIR conference on Research and Development in Information Retrieval. SIGIR ’16: The 39th International ACM SIGIR conference on research and development in Information Retrieval. ACM. https://doi.org/10.1145/2911451.2914726
19. Vasile, F., Smirnova, E., & Conneau, A. (2016). Meta-Prod2Vec. Içinde Proceedings of the 10th ACM Conference on Recommender Systems. RecSys ’16: Tenth ACM Conference on Recommender Systems. ACM. https://doi.org/10.1145/2959100.2959160
20. Grbovic, M., Radosavljevic, V., Djuric, N., Bhamidipati, N., Savla, J., Bhagwan, V., & Sharp, D. (2015). E-commerce in Your Inbox. Içinde Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. KDD ’15: The 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. ACM. https://doi.org/10.1145/2783258.2788627
21. Hsieh, C.-K., Yang, L., Cui, Y., Lin, T.-Y., Belongie, S., & Estrin, D. (2017). Collaborative Metric Learning. Içinde Proceedings of the 26th International Conference on World Wide Web. WWW ’17: 26th International World Wide Web Conference. International World Wide Web Conferences Steering Committee. https://doi.org/10.1145/3038912.3052639
22. Roy, S., & Guntuku, S. C. (2016). Latent Factor Representations for Cold-Start Video Recommendation. Içinde Proceedings of the 10th ACM Conference on Recommender Systems. RecSys ’16: Tenth ACM Conference on Recommender Systems. ACM. https://doi.org/10.1145/2959100.2959172
23. Li, S., Kawale, J., & Fu, Y. (2015). Deep Collaborative Filtering via Marginalized Denoising Auto-encoder. Içinde Proceedings of the 24th ACM International on Conference on Information and Knowledge Management. CIKM’15: 24th ACM International Conference on Information and Knowledge Management. ACM. https://doi.org/10.1145/2806416.2806527
24. Zheng, L., Noroozi, V., & Yu, P. S. (2017). Joint Deep Modeling of Users and Items Using Reviews for Recommendation. Içinde Proceedings of the Tenth ACM International Conference on Web Search and Data Mining. WSDM 2017: Tenth ACM International Conference on Web Search and Data Mining. ACM. https://doi.org/10.1145/3018661.3018665
25. Bansal, T., Belanger, D., & McCallum, A. (2016). Ask the GRU: Multi-task Learning for Deep Text Recommendations. Içinde Proceedings of the 10th ACM Conference on Recommender Systems. RecSys ’16: Tenth ACM Conference on Recommender Systems. ACM. https://doi.org/10.1145/2959100.2959180
26. Zhang, S., Yao, L., Sun, A., & Tay, Y. (2020). Deep learning based recommender system: A survey and new perspectives. ACM Computing Surveys, 52(1), 1–38. https://doi.org/10.1145/3285029
27. Wu, Y., DuBois, C., Zheng, A. X., & Ester, M. (2016). Collaborative Denoising Auto-Encoders for Top-N Recommender Systems. Içinde Proceedings of the Ninth ACM International Conference on Web Search and Data Mining. WSDM 2016: Ninth ACM International Conference on Web Search and Data Mining. ACM. https://doi.org/10.1145/2835776.2835837
28. Wang, J., Kenthapadi, K., Rangadurai, K., & Hardtke, D. (2017). Dionysius: A Framework for Modeling Hierarchical User Interactions in Recommender Systems (Versiyon 1). arXiv. https://doi.org/10.48550/ARXIV.1706.03849
29. Zhang, S., Yao, L., Sun, A., & Tay, Y. (2019). Deep Learning Based Recommender System. Içinde ACM Computing Surveys (C. 52, Issue 1, ss. 1-38). Association for Computing Machinery (ACM). https://doi.org/10.1145/3285029
30. Wang, M. (2020). Applying Internet information technology combined with deep learning to tourism collaborative recommendation system. Içinde Z. Lv (Ed.), PLOS ONE (C. 15, Issue 12, s. e0240656). Public Library of Science (PLoS). https://doi.org/10.1371/journal.pone.0240656
31. Çakır, M., Öğüdücü, Ş. G., & Tugay, R. (2019). A Deep Hybrid Model for Recommendation Systems. Içinde Lecture Notes in Computer Science (ss. 321-335). Springer International Publishing. https://doi.org/10.1007/978-3-030-35166-3_23
32. Covington, P., Adams, J., & Sargin, E. (2016). Deep Neural Networks for YouTube Recommendations. Içinde Proceedings of the 10th ACM Conference on Recommender Systems. RecSys ’16: Tenth ACM Conference on Recommender Systems. ACM. https://doi.org/10.1145/2959100.2959190
33. Geron, A. (2019). Hands-on machine learning with scikit-learn, keras, and TensorFlow: Concepts, tools, and techniques to build intelligent systems (2nd ed.). O’Reilly Media.
34. Chollet, F. (2022). Deep learning with python. Manning Publications.
35. Goodfellow, I., Bengio, Y., & Courville, A. (2016). Deep Learning. MIT Press.
36. Hastie, T., Tibshirani, R., & Friedman, J. (2009). The Elements of Statistical Learning: Data Mining, Inference, and Prediction. Springer Science & Business Media.
37. Cram101 Textbook Reviews. (2013). Studyguide for pattern recognition and machine learning by bishop, Christopher M. Cram101.
38. Ramachandran, P., Zoph, B., & Le, Q. V. (2017). Searching for Activation Functions (Versiyon 2). arXiv. https://doi.org/10.48550/ARXIV.1710.05941
39. Rumelhart, D. E., Hinton, G. E., & Williams, R. J. (1986). Learning representations by back-propagating errors. Nature, 323(6088), 533–536. https://doi.org/10.1038/323533a
40. Bottou, L. (2010). Large-Scale Machine Learning with Stochastic Gradient Descent. Içinde Proceedings of COMPSTAT’2010 (ss. 177-186). Physica-Verlag HD. https://doi.org/10.1007/978-3-7908-2604-3_1
41. Ruder, S. (2016). An overview of gradient descent optimization algorithms (Versiyon 2). arXiv. https://doi.org/10.48550/ARXIV.1609.04747
42. Sun, R.-Y. (2020). Optimization for Deep Learning: An Overview. Içinde Journal of the Operations Research Society of China (C. 8, Issue 2, ss. 249-294). Springer Science and Business Media LLC. https://doi.org/10.1007/s40305-020-00309-6
43. Ioffe, S., & Szegedy, C. (2015). Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift (Versiyon 3). arXiv. https://doi.org/10.48550/ARXIV.1502.03167
44. Jose, G. V. (2019, February 10). Effect of Learning rate on Loss. Towards Data Science. https://towardsdatascience.com/useful-plots-to-diagnose-your-neural-network-521907fa2f45
45. Jose, G. V. (2019, February 10). Accuracy Plot. Towards Data Science. https://towardsdatascience.com/useful-plots-to-diagnose-your-neural-network-521907fa2f45
46. Lecun, Y., Bottou, L., Bengio, Y., & Haffner, P. (1998). Gradient-based learning applied to document recognition. IEEE. Institute of Electrical and Electronics Engineers, 86(11), 2278–2324. https://doi.org/10.1109/5.726791
47. Hinton, G. E. (2012). A Practical Guide to Training Restricted Boltzmann Machines. Içinde Lecture Notes in Computer Science (ss. 599-619). Springer Berlin Heidelberg. https://doi.org/10.1007/978-3-642-35289-8_32
48. N. Srivastava, G. Hinton, A. Krizhevsky, I. Sutskever, and R. Salakhutdinov, Dropout: a simple way to prevent neural networks from overfitting, The journal of machine learning research, vol. 15, no. 1, pp. 1929–1958, 2014.
49. Taylor, P. (2022, September 8). Amount of data created, consumed, and stored 2010-2020, with forecasts to 2025. Statista. https://www.statista.com/statistics/871513/worldwide-data-created/
50. Uzun-Per, M., Can, A. B., Volkan Gurel, A., & Aktas, M. S. (2021). Big Data Testing Framework for Recommendation Systems in e-Science and e-Commerce Domains. Içinde 2021 IEEE International Conference on Big Data (Big Data). 2021 IEEE International Conference on Big Data (Big Data). IEEE. https://doi.org/10.1109/bigdata52589.2021.9672082
51. Uzun‐Per, M., Gurel, A. V., Can, A. B., & Aktas, M. S. (2022). Scalable recommendation systems based on finding similar items and sequences. Içinde Concurrency and Computation: Practice and Experience (C. 34, Issue 20). Wiley. https://doi.org/10.1002/cpe.6841
52. Sancar, S., & Uzun-Per, M. (2023). Testing the Performance of Feature Selection Methods for Customer Churn Analysis: Case Study in B2B Business. Içinde Computational Intelligence, Data Analytics and Applications (ss. 509-519). Springer International Publishing. https://doi.org/10.1007/978-3-031-27099-4_39
53. Sancar, S., & Uzun-Per, M. (2022). Feature Selection in Customer Churn Analysis: Case Study in B2B Business. Içinde 2022 IEEE International Conference on e-Business Engineering (ICEBE). 2022 IEEE International Conference on e-Business Engineering (ICEBE). IEEE. https://doi.org/10.1109/icebe55470.2022.00053