Trip planning based on sequential recommender systems using textual  representations

dc.contributor.author	Yatskiv, Bohdan
dc.date.accessioned	2024-02-14T13:18:21Z
dc.date.available	2024-02-14T13:18:21Z
dc.date.issued	2022
dc.identifier.citation	Yatskiv, Bohdan. Trip planning based on sequential recommender systems using textual representations / Yatskiv, Bohdan; Supervisor: Taras Firman; Ukrainian Catholic University, Department of Computer Sciences. – Lviv: 2022. – 34 p.	uk
dc.identifier.uri	https://er.ucu.edu.ua/handle/1/4455
dc.language.iso	en	uk
dc.title	Trip planning based on sequential recommender systems using textual representations	uk
dc.type	Preprint	uk
dc.status	Публікується вперше	uk
dc.description.abstracten	Finding interesting places to visit is one of the most common problems while travel- ling. With the development of review services, a large amount of personalized infor- mation is produced by the users. We believe that the information that review texts contain could be used to improve the personalized recommendations. This work fo- cuses on creating a sequential recommender system that provides recommendations based on the chronological history of users’ text reviews. The proposed model uses BERT text embeddings to get review text representations and a Transformer encoder to learn the context between items in sequence with the multi-attention mechanism. This approach allows to provide relevant recommendations based on the user’s se- quential behavior and makes the trip planning more comfortable.	uk