First Master's thesis abstract (Towards Fully Automated Price Comparison: A Similarity-Based Approach to Product Labeling)

With the rise of e-commerce, users increasingly rely on price comparison platforms to evaluate products and prices across multiple retailers. These platforms depend on two core components: web crawling (extracting product data) and product labeling (grouping similar items). While crawling is often automated, labeling typically involves manual effort. This thesis introduces an automated labeling method using text similarity algorithms—such as Jaccard and Levenshtein distances—to cluster semantically similar product titles and automatically assign a representative title to each group. To assess the effectiveness of this approach, a mathematical evaluation metric is also proposed to quantify labeling accuracy. By automating both product grouping and representative title selection, and by providing a rigorous evaluation framework, this work contributes to building more scalable and intelligent comparison systems.

Keywords: Web Crawling, Price Comparison, Product Labeling, Text Similarity, Clustering, Evaluation Metric

Bachelor project explainations (Implementation of an e-commerce website like ebay.com)

This project was an implementation of a B2B E-commerce website that I did in about 2013, When I was only 23 years old. I worked at a web developing company, and I was an expert in designing and programming at that age. For both, designing(visually) and programming, I did not use any prepared Content Management System, and the website fully developed by myself from the base, and without seeking help from anyone. The website had a lot of parts and functionality. You can see the picture of the home page below: