Item2Vec lähenemine soovitussüsteemi jaoks
Laen...
Failid
Kuupäev
Autorid
Ajakirja pealkiri
Ajakirja ISSN
Köite pealkiri
Kirjastaja
Abstrakt
Lõputöö eesmärk oli välja arendada sõnade vektoriesitusel põhinev mudel soovitussüsteemi jaoks. Mudelit rakendati Item2Vec meetodi alusel, kuna viimased uuringud on näidanud, et Item2Veci abil on võimalik saada konkurentsivõimelisi tulemusi. Töö koosnebki valitud lähenemise rakendamisest, selle tarkvara tehnoloogia paigaldamise kirjeldamisest Tarkvara Tehnoloogia Arenduskeskuse (STACC) tootele ning testandmete hindamise ülevaatest. Hindamise tulemused näitavad, et rakendatud mudel tõepoolest sobib soovitussüsteemide arendamiseks.
The aim of this thesis was to develop a vector space representation model for a recommender system. The model implementation was based on the method called Item2Vec; this approach was chosen because recent studies have shown that it displays competitive results. The work consists of implementing the approach, evaluating it on test data and deploying it into production in the Software Technology and Applications Competence Center. Evaluation results show that the implemented model is indeed competitive and is suitable for building recommender systems.
The aim of this thesis was to develop a vector space representation model for a recommender system. The model implementation was based on the method called Item2Vec; this approach was chosen because recent studies have shown that it displays competitive results. The work consists of implementing the approach, evaluating it on test data and deploying it into production in the Software Technology and Applications Competence Center. Evaluation results show that the implemented model is indeed competitive and is suitable for building recommender systems.