POLIBITS, Vol. 57, pp. 81-88, 2018.
Abstract: information extraction, lightweight ontology, natural language processing, recommender systems
Keywords: In this paper a recommender system is described which takes a set of venue categories of user’s interest into ac- count to form a tourist itinerary throughout a city. The system is focused on user preferences in venue aspects. Techniques of such aspects extraction are developed in this paper, in particular from reviews corpora. User preferences are used to weigh aspects associated with particular sights and restaurants. These filtered venues along with time restrictions are subject to submit into the recommender system. A lightweight ontology is discussed which describes the domains of restaurants and sightseeing knowledge and allows venues comparative analysis to enhance the search for relevant venues. The system designed performs automated planning of tourist itineraries, flexible sights searching, and analysis of venues aspects extracted from reviews in Russian.
PDF: Recommender System for Tourist Itineraries Based on Aspects Extraction from Reviews Corpora
PDF: Recommender System for Tourist Itineraries Based on Aspects Extraction from Reviews Corpora
https://doi.org/10.17562/PB-57-9
Table of contents of POLIBITS 57