Using Machine Learning to Detect if Two Products Are the Same
Využití strojového učení pro detekování, kdy jsou dva produkty stejné
Authors
Supervisors
Reviewers
Editors
Other contributors
Journal Title
Journal ISSN
Volume Title
Publisher
České vysoké učení technické v Praze
Czech Technical University in Prague
Czech Technical University in Prague
Date
Abstract
V tejto práci sa zamieravame na možnosti využitia strojového učenia v oblasti e-commerce. S konkrétnym využitím pre párovanie produktov a ich ponúk od roznych obchodov. Aj ked všetky metódy budú optimalizované pre toto použitie, ich techniky sa mozu neskor využiť aj na iné oblasti, ako napríklad obohacovanie katalógu produktov o nové parametre pre produkty alebo pokročilé formy vyhľadávania. V závere využijeme naprogramované REST API, ktoré využíva náš model, na evaluáciu nad reálnymi problémami, ktoré postihujú dnešné online katalógy produktov. A to zamezenie duplicitám a zle napárovaných ponúk od obchodov k produktom.
In this work, we investigate ways to use machine learning in the e-commerce field, with an application for the problem of pairing different descriptions of the same product from various online shops. Even though we evaluate the methods developed in this thesis only on this problem, they could be used in various areas. In addition, we create a new REST API and use it to evaluate our model on real-world datasets. Specifically, we apply our methods for finding duplicates in an existing online catalog aggregating items from hundreds of e-shops.
In this work, we investigate ways to use machine learning in the e-commerce field, with an application for the problem of pairing different descriptions of the same product from various online shops. Even though we evaluate the methods developed in this thesis only on this problem, they could be used in various areas. In addition, we create a new REST API and use it to evaluate our model on real-world datasets. Specifically, we apply our methods for finding duplicates in an existing online catalog aggregating items from hundreds of e-shops.