Perakendecilikte ikame ürün seçimi için Gower benzerliği tabanlı bir yaklaşım

Öz

Bu çalışma, perakende sektöründe müşteri memnuniyetini artırma ve satış performansını optimize etmede ürün ikamesinin önemli rolünü ele almaktadır. İkame, istenen bir ürünün mevcut olmaması durumunda gerçekleşir ve tüketicilerin benzer ihtiyaçları karşılayan alternatif ürünleri seçmesine yol açar. Bu süreç, özellikle geniş ürün çeşitliliğine ve yüksek müşteri beklentilerine sahip sektörlerde kritik öneme sahiptir. Çalışma, geçmiş satış verileri olmasa bile, yalnızca ürün özelliklerine dayanarak ikame ürünleri belirlemek için etkili bir yöntem olarak Gower benzerlik katsayısını tanıtmaktadır. Gower benzerlik metriğinden yararlanan yaklaşım, ürün benzerliklerini hesaplamak için kategorik, ikili ve sürekli değişkenler gibi karma veri türlerini entegre eder. Sonuçlar, önerilen yöntemin satış davranışında yüksek tutarlılığa sahip ikameleri belirlediğini ve böylece bu veri odaklı çerçevenin güvenilirliğini doğruladığını göstermektedir. Bulgular, talep tahmini ve ürün çeşitliliği planlamasına değerli içgörüler sunarak, dinamik perakende pazarlarında ürün bulunmamasıyla ilgili riskleri azaltmak ve envanter kararlarını optimize etmek için etkili bir çözüm sunmaktadır.

Anahtar Kelimeler

• ürün ikamesi
• veri odaklı perakende optimizasyonu
• gower benzerliği
• talep tahmini
• envanter yönetimi

Gower similarity-based approach for substitution product selection in retailing

Öz

This study addresses the significant role of product substitution in retail for enhancing customer satisfaction and optimizing sales performance. Substitution occurs when a desired product is unavailable, leading consumers to select alternative products that fulfill similar needs. This process is particularly crucial in industries with extensive product variety and high customer expectations. The study introduces Gower's Similarity Score as an effective method for identifying substitute products based solely on product attributes, even in the absence of historical sales data. By leveraging Gower's similarity metric, the approach integrates mixed data types—categorical, binary, and continuous variables—to calculate product similarities. The results demonstrate that the proposed method identifies substitutes with high consistency in sales behavior, thereby validating the reliability of this data-driven framework. The findings contribute valuable insights into demand forecasting and assortment planning, offering a robust solution for mitigating risks related to product unavailability and optimizing inventory decisions in dynamic retail markets.

Anahtar Kelimeler

• product substitution
• data-driven retail optimization
• gower similarity
• demand forecasting
• inventory management

Kaynakça

1. 1. Tkachuk S., Wróblewska A., Dabrowski J., Lukasik S., Identifying Substitute and Complementary Products for Assortment Optimization with Cleora Embeddings, 2022 International Joint Conference on Neural Networks (IJCNN), 1–7, 2022.
2. 2. Das N., Joshi A., Yenigalla P., Agarwal G., MAPS: Multimodal Attention for Product Similarity, 2022 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), 2988–2996, 2022.
3. 3. Zuo Z., Wang L., Momma M., Wang W., Ni Y., Lin J.S., A Flexible Large-Scale Similar Product Identification System in E-Commerce, 2020.
4. 4. Tüzün Aksu D., Durak B., Civelek D., Tortop S.S., A Decomposition-Based Solution for Shelf Space Allocation and Assortment Optimization with Space Elasticity, Substitution Effects and Multiple Facing Options for the Retail Sector, Journal of the Faculty of Engineering and Architecture of Gazi University, 40 (4), 2381–2391, 2025.
5. 5. Göçen G., Aksoy A., Integrated Management System for Vending Machines, Journal of the Faculty of Engineering and Architecture of Gazi University, 39 (3), 1893–1906, 2024.
6. 6. Gopal P., Mudunuri S., Dutta S., Motwani K., Uncovering Critical Products in Retail Baskets: A Predictive Modelling Approach to Increase Order Fulfilment, Proceedings of the Third International Conference on AI-ML Systems, 2023.
7. 7. Akkurt T., Sarıçiçek İ., KPI Based Performance Estimation in Production Systems Using Deep Learning Techniques, Journal of the Faculty of Engineering and Architecture of Gazi University, 39 (3), 1499–1507, 2024.
8. 8. Pande A., Gupta A., Ni K., Biswas R., Majumdar S., Substitution Techniques for Grocery Fulfillment and Assortment Optimization Using Product Graphs, 2020.

9. 9. Wang W., Cui Y., Li G., Jiang C., Deng S., A Self-Attention-Based Destruction and Construction Learning Fine-Grained Image Classification Method for Retail Product Recognition, Neural Computing and Applications, 32, 14613–14622, 2020.
10. 10. Kenardi M.P., The S., Rahmania R., Self-Attention Approach for Inter-Class Similarities of Grocery Product Classification, 7th International Conference on Informatics and Computational Sciences (ICICoS), 179–184, 2024.
11. 11. Li J., Dou Z., Zhu Y., Zuo X., Wen J.R., Deep Cross-Platform Product Matching in E-Commerce, Information Retrieval Journal, 23, 136–158, 2019.
12. 12. Tian Y., Lautz S., Wallis A., Lambiotte R., Extracting Complements and Substitutes from Sales Data: A Network Perspective, EPJ Data Science, 10, 2021.
13. 13. Derhami S., Montreuil B., Estimation of Potential Lost Sales in Retail Networks of High-Value Substitutable Products, IISE Transactions, 54, 563–577, 2022.
14. 14. Zhang M., Wei X., Guo X., Chen G., Wei Q., Identifying Complements and Substitutes of Products, ACM Transactions on Knowledge Discovery from Data, 13, 1–29, 2019.
15. 15. Koren M., Perlman Y., Shnaiderman M., Inventory Management for Stockout-Based Substitutable Products Under Centralised and Competitive Settings, International Journal of Production Research, 62, 3176–3192, 2023.
16. 16. Gupta V., Ivanov D., Choi T., Competitive Pricing of Substitute Products Under Supply Disruption, Omega, 101, 102279, 2020.
17. 17. Zhuravlev Y.I., Dokukin A., Senko O., Stefanovskiy D., Use of Clusterization Technique to Highlight Groups of Related Goods by Digital Traces in Retail Trade, 9th International Conference on Advanced Computer Information Technologies (ACIT), 84–88, 2019.
18. 18. Zhang X., Dearden J.A., Yao Y., Let Them Stay or Let Them Go? Online Retailer Pricing Strategy for Managing Stockouts, Production and Operations Management, 31, 4173–4190, 2022.
19. 19. Gower J.C., A General Coefficient of Similarity and Some of Its Properties, Biometrics, 27 (4), 857–871, 1971.
20. 20. Santos T.R.L., Zárate L.E., Categorical Data Clustering: What Similarity Measure to Recommend?, Expert Systems with Applications, 42, 1247–1260, 2015.
21. 21. Mahara T., Sharma S., Kumar A., Sangaiah A.K., A Cognitive Similarity-Based Measure to Enhance the Performance of Collaborative Filtering-Based Recommendation System, IEEE Transactions on Computational Social Systems, 9, 1785–1793, 2022.
22. 22. Buyaktif, DAC’22 – Invent Analytics Project, Kaggle, https://kaggle.com/competitions/dac22-invent-analytics-project, Yayın tarihi 2022, Erişim tarihi 2026.
23. 23. Akkurt N., Hasgül S., Comparison of Automated Machine Learning (AutoML) Libraries in Time Series Forecasting, Journal of the Faculty of Engineering and Architecture of Gazi University, 39 (3), 1693–1701, 2024.
24. 24. Wan M., Huang Y., Zhao L., Deng T., Fransoo J.C., Demand Estimation Under Multi-Store Multi-Product Substitution in High Density Traditional Retail, European Journal of Operational Research, 266 (1), 99–111, 2018.
25. 25. Hwangbo H., Kim Y.S., Cha K., Recommendation System Development for Fashion Retail E-Commerce, Electronic Commerce Research and Applications, 28, 94–101, 2018.
26. 26. Hoang D., Breugelmans E., “Sorry, the Product You Ordered Is Out of Stock”: Effects of Substitution Policy in Online Grocery Retailing, Journal of Retailing, 99 (1), 26–45, 2023.