Investigation of the influence of textual representation of goods by means of models using artificial neural networks of deep learning on the relevance of the search for goods on the electronic trading Internet platform

Fedor Krasnov


With the growing popularity of online shopping, academic research in the field of e-commerce is becoming increasingly relevant. Nevertheless, significant research problems remain, starting with classical search problems in e-commerce, for example, the comparison of text queries with multimodal representations of goods, optimization of ranking for bilateral Internet trading platforms, taking into account the interaction of buyer and seller with recommendation and search engines. These areas of research are important for understanding customer behavior, stimulating engagement, and increasing conversions. Product data plays a key role in search and recommendations. Services in which there are several suppliers (marketplaces) on the electronic trading Internet platform demonstrate high dynamics in terms of the quality and consistency of content, fraud detection and pricing. Inventory updates are often accompanied by service level agreements that guarantee that changes will be made in accordance with customer requirements within a strictly defined time frame. Finally, the goods themselves can be considered as multimodal documents, and the success of the search depends on the compliance of the client's intentions with all aspects of the goods. For the sake of specificity, we will use the term "documents" for the elements returned by the search query, although the returned elements may be more general (for example, multimedia elements).

When searching on an online electronic trading platform, the indexed products that buyers are looking for are combinations of images, videos and unstructured text (names, descriptions and reviews) and structured data (price, brand, ratings, seller's location, and logistics). This combination of data opens up prospects for research, including improving the extraction of data from a variety of sources using signals from different types of data, such as providing more detailed color information and using image similarity for recommendations. and also as a way for customers to create queries in a search engine. Data is the key to providing high-quality search and recommendations that meet the needs of customers and businesses. In this article, the author explored the possibilities of several approaches to extracting useful signals from various sources of product information to improve the customer experience in the field of e-commerce.

Full Text:

PDF (Russian)


Yang Y. Query-aware tip generation for vertical search // Proceedings of the 29th ACM International Conference on Information & Knowledge Management. – 2020. – P. 2893-2900.

Ibrahim, Osman Ali Sadek and Eman M. G. Younis. Hybrid online–offline learning to rank using simulated annealing strategy based on dependent click model // Knowledge and Information Systems. – 2022. – Vol. 64. – P. 2833-2847.

Lyu, Zequn Deep Match to Rank Model for Personalized Click-Through Rate Prediction // AAAI Conference on Artificial Intelligence. – 2020.

Zhai, Jiaqi Revisiting Neural Retrieval on Accelerators // ArXiv abs/2306.04039. – 2023.

C. Rudin. The p-norm push: A simple convex ranking algorithm that concentrates at the top of the list // The Journal of Machine Learning Research. – 2009. – Vol. 10. – P. 2233–2271.

Krauth, K., Dean, S., Zhao, A., Guo, W., Curmei, M., Recht, B., & Jordan, M. I. Do offline metrics predict online performance in recommender systems? // arXiv preprint arXiv:2011.07931. – 2020.

Curmei, M., Haupt, A. A., Recht, B., & Hadfield-Menell, D. Towards psychologically-grounded dynamic preference models // In Proceedings of the 16th ACM Conference on Recommender Systems. – 2022. – P. 35-48.

Patro, G. K., Porcaro, L., Mitchell, L., Zhang, Q., Zehlike, M., & Garg, N. Fair ranking: a critical review, challenges, and future directions // In Proceedings of the 2022 ACM Conference on Fairness, Accountability, and Transparency. – 2022 (June). – P. 1929-1942.

Olivier Chapelle, Donald Metlzer, Ya Zhang, and Pierre Grinspan Expected reciprocal rank for graded relevance // In Proceedings of the 18th ACM conference on Information and knowledge management. ACM. – 2009. – P. 621–630.

Yun Zhou and W Bruce Cro. Query performance prediction in web search environments // In Proceedings of the 30th annual international ACM SIGIR conference on Research and development in information retrieval. ACM. – 2007. – P. 543–550.

Cronen-Townsend, Steve [et al] Predicting query performance // Annual International ACM SIGIR Conference on Research and Development in Information Retrieva. – 2002.

Järvelin, Kalervo and Jaana Kekäläinen Cumulated gain-based evaluation of IR techniques // ACM Trans. Inf. Syst. – 2002. – Vol. 20. – P. 422-446.

Evangelos Kanoulas and Javed A. Aslam. Empirical justification of the gain and discount. – 2009.

Carmel, D., Haramaty, E., Lazerson, A., & Lewin-Eytan, L. Multi-objective ranking optimization for product search using stochastic label aggregation // In Proceedings of The Web Conference. – 2020 (April). – P. 373-383.

Hong, Liangjie, and Mounia Lalmas Tutorial on online user engagement: Metrics and optimization // Companion Proceedings of The 2019 World Wide Web Conference. – 2019.

Reddy, Chandan K. Shopping queries dataset: A large-scale ESCI benchmark for improving product search // arXiv preprint arXiv:2206.06588. – 2022.

Tsagkias, Manos, Wouter Weerkamp, and Maarten De Rijke News comments: Exploring, modeling, and online prediction // Advances in Information Retrieval: 32nd European Conference on IR Research, ECIR. – Milton Keynes, UK, March 28-31, 2010.

Bojanowski P. Enriching word vectors with subword information // Transactions of the association for computational linguistics. – 2017. – Vol. 5. – P. 135-146.

Rogers A. RuSentiment: An enriched sentiment analysis dataset for social media in Russian // Proceedings of the 27th international conference on computational linguistics. – 2018. – P. 755-763.

Huang P. S. Learning deep structured semantic models for web search using clickthrough data // Proceedings of the 22nd ACM international conference on Information & Knowledge Management. – 2013. – P. 2333-2338.


  • There are currently no refbacks.

Abava  Кибербезопасность MoNeTec 2024

ISSN: 2307-8162