Meno: | Patrícia
|
---|
Priezvisko: | Vnenčáková
|
---|
Názov: | Course recommender for FMPH CU students
|
---|
Vedúci: | Mgr. Askar Gafurov, PhD.
|
---|
Rok: | 2023
|
---|
Kµúčové slová: | recommender system, model, collaborative filtering, similarity method, evaluator, accuracy metric, data
|
---|
Abstrakt: | Modern targeted marketing uses a wide palette of methods to determine the best offer of products to individual customers. Those methods are collectively called recommender systems. The goal of this thesis is an application of such methods in order to recom- mend courses to university students. In this thesis, we implemented a recommender system based on the k-NN algorithm with three similarity methods. We evaluated our recommender on anonymized data of students of FMPH CU from the academic years 2009/10 to 2021/22. The results showed that the best similarity method is the Jaccard index. A considerable part of this thesis was data cleaning, such as the merging of multiple versions of courses that nevertheless represented the same course.
|
---|