Log-Based Session Profiling and Online Behavioral Prediction in E–Commerce Websites

Abstract

Improvements to customer experience give companies a competitive advantage, as understanding customers' behaviors allows e-commerce companies to enhance their marketing strategies by means of recommendation techniques and the customization of products and services. This is not a simple task, and it becomes more difficult when working with anonymous sessions since no historical information of the user can be applied. In this paper, analysis and clustering of the clickstreams of past anonymous sessions are used to synthesize a prediction model based on a neural network. The model allows for prediction of a user's profile after a few clicks of an online anonymous session. This information can be used by the e-commerce's decision system to generate online recommendations and better adapt the offered services to the customer's profile. INDEX TERMS Behavior prediction, User profiling, Log analysis, Clustering, Neural networks, Model checking.

Improving Ads-Profitability Using Traffic-Fingerprints

“…For example, in [10] users are being segmented based on their social media advertising perceptions which reveals groups of users more susceptible to such ads. Another context for clustering users based on behaviour is web page recommendation [5] or product recommendation [17]. Although these papers use contextual data, some approaches create user profiles solely based on click-stream data [13,11,14,3,15,7].…”

“…For determining the optimal number of clusters, for each ad network we start with 2 clusters and then gradually increase the number as far as adding another cluster does not give a relevant improvement of a sum of differences between elements and clusters' centers, according to the so-called Elbow method 5 .…”

The Impact of Social Media Technologies on Supplier and Customer Relationship Management: An Empirical Analysis in the U.S.

“…Consequently, it is not possible to observe the specific process of how relationships with suppliers and customers evolve as social networks are formed through social media channels. In particular, as the importance of non-face-to-face communication has rapidly increased due to the recent COVID-19 pandemic, understanding how the rise of such non-face-to-face manner has changed the role of social media in supply chain relationship management is expected to be an essential research topic [96], [97]. Therefore, we expect that future research will find more impressive results by using longitudinal data to explore the evolution of social media effects on CRM and SRM implementations.…”

Multi-Criteria Group Assessment of E-Commerce Websites Based on the New PROSA GDSS Method–the Case of Poland

“…On the other hand, the benefits for sellers are certainly: minimizing operating costs, increasing business efficiency and improving the market reach [10]. Nevertheless, in the 2019-2020 period, the main reason for the growth of the e-commerce market was the fact that in many countries during subsequent lockdowns, it was the only possible way for customers to purchase various goods [12]. On the other hand, from the sellers' perspective, it should be noted that the transition from stationary sales to ecommerce allowed many of them to survive the lockdowns [8].…”