Mapping the Evolution of Algorithmic Recommendation Systems in E-Commerce (2020–2025): Productivity, Impact, and Thematic Hotspots from a Bibliometric Study

Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).

Ecommerce platforms use recommendation systems to make things personal for users andincrease sales. These systems are really changing the way people shop online. In fact,ecommerce sales are going to reach $6.4 trillion by 2025, which is more than 21% of all retailsales worldwide (eMarketer, Global ecommerce forecast 2024). This means we needrecommendation systems that grab people’s attention and help them decide what to buy.However, with all this growth there is not a lot of research that focuses on ecommercerecommendation systems. Most research is broader. Looks at ecommerce in general orrecommendation systems in other industries. This paper wants to fill that gap by looking at146 research articles from 2020 to 2025 that were found in the Scopus database.The research used a tool called Bibliometrix to look at things like how many times articleswere cited, who wrote them and what keywords were used. It found that 485 authors wrotethese articles and used 1,179 keywords, which shows how many different fields are involvedin this research. The results show that more and more research is being done in this area in2023 and 2024 because of the increase in online shopping.The most talked about topics are filtering, deep learning, personalization and using artificialintelligence for recommendations. The research also found that authors from countries andfields are working together which shows how international and interdisciplinary this researchis. Some authors and papers are cited a lot, which means they are important in the field. Theyoften combine computer science, information systems and marketing.This study gives us a look at what is happening in ecommerce recommendation research andcan help researchers, people who work in the field and those who build platforms make thingsbetter for users. Ecommerce recommendation systems are a part of this and this research canhelp us understand them better. The goal is to make things more personal, for users andimprove their experience. Ecommerce platforms and recommendation systems are. Thisresearch can help us see how they work together..

• Optimization of Machine Learning–Based Recommendation Systems on E-Commerce Platforms

  East Asian Journal of Multidisciplinary Research 2026 · 0 citations