| Title |
Deep neural network-based personalized book recommender system feasibility study for libis / |
| Translation of Title |
Giliaisiais neuroniniais tinklais grįstos suasmenintos knygų rekomendacijų sistemos įgyvendinamumo tyrimas LIBIS sistemoje. |
| Authors |
Senvaitis, Kazimieras |
| Full Text |
|
| Pages |
64 |
| Keywords [eng] |
rekomendacinės sistemos, viešosios bibliotekos, LIBIS, METIS, gilieji neuroniniai tinklai, bendradarbiavimo filtravimas, demografinis filtravimas, turiniu grįstas filtravimas, hibridinis modelis, recommender system, public library, LIBIS, METIS, deep neural networks, collaborative filtering, demographics filtering, content-based filtering, hybrid model |
| Abstract [eng] |
Recommender systems (RSs) are becoming an inseparable part of our everyday lives. In vast volumes of data, they help us find our favorite items to purchase, our friends on social networks, and our favorite movies to watch. However, in the domain of public libraries, RSs are yet poorly covered even though RS science field is currently flourishing due to the advancements in machine learning and high business value. This master thesis bridges this gap by offering in-depth feasibility study of implementing deep neural network-based personalized book recommender system for Lithuanian public library software LIBIS with emphasis on privacy preservation. In this paper, LIBIS data and limitations regarding book orders are presented and a dataset of 20 million book orders has been developed. Multiple industry-leading collaborative filtering (CF) algorithms have been evaluated on the aforementioned dataset. Furthermore, a hybrid model has been developed by combining best-performing variant of CF with Demographics filtering model and Content-based filtering model into a robust recommender system, which is capable of resolving new user cold-start and new item cold-start problems, while for users with long reading history provides diverse, novel, serendipitous recommendations. Lastly, integration of the RS into LIBIS system is presented. |
| Dissertation Institution |
Vilniaus universitetas. |
| Type |
Master thesis |
| Language |
English |
| Publication date |
2022 |
