Extending collaborative filtering recommendation using word embedding: A hybrid approach

Abstract

Summary Collaborative filtering recommendation systems, which analyze sets of user ratings, have been applied to various domains and have resulted in considerable improvements in the traditional recommendation system. However, they still have problems with data sparsity and cold‐start of the user ratings. To solve these problems, we present a hybrid recommendation approach by combining collaborative filtering methods and word embedding‐based content analysis. This study focuses on the movie domain, and therefore, the contents of the items are represented as a set of features such as titles, genres, directors, actors, and plots. The main aim of this paper is to understand the content of the movie plot using a word embedding to improve the measurement of similarity of each plot content to other plot content (called plot embedding). To enhance the accuracy in measuring the similarity between movies, we also consider other features such as titles, genres, directors, and actors extracted from movies. In the experiments, the movie dataset was collected by our crowdsourcing platform, which is the OMS platform. The experimental findings indicate that the proposed approach can enhance the efficiency of applied collaborative filtering recommendation systems.

Recommendation Algorithm Using Clustering-Based UPCSim (CB-UPCSim)

“…Recently, some studies [36][37][38] also proposed sophisticated collaborative filtering to improve recommendation performances. The study conducted by [36] suggested new collaborative filtering using cognitive similarity.…”

“…We still need further study to compare both algorithms for higher k (i.e., k = 60 up to k = 100) because the cognitive similarity did not measure them. In addition, Nguyen et al [37] use word embedding to improve their proposed collaborative filtering. Hence, implementing word embedding in CB-UPCSim can be another option to obtain better performance.…”

Multi Sequential Embedding-based Cross-Domain Recommendation

“…• ItemPop, as a non-personalized method, ranks items according to their popularity measured by their interaction numbers. • KIU proposed by [35] is based on the similarity trained by Word2vec to recommend items to users. To obtain user and item representation vectors, it constructs sentences that start with a user identity followed by the identity sequence of the items chronologically consumed by the user.…”

DaGzang: a synthetic data generator for cross-domain recommendation services

“…Recommendation services aim to model user preferences based on user history interactions, such as item ratings ( Nguyen et al, 2020a ; Hong & Jung, 2018 ; Nguyen, Jung & Hwang, 2020b ; Vuong Nguyen et al, 2021 ). Some of the traditional methods in recommendation systems (RSs) such as matrix factorization (MF) ( Koren, RM & Volinsky, 2009 ; Salakhutdinov & Mnih, 2007 ) or neural collaborative factoring (CF) ( Cheng et al, 2016 ; Dziugaite & Roy, 2015 ; He et al, 2017 ; Nguyen, Nguyen & Jung, 2020c ) are applied to several specific datasets collected from several sources.…”

“…The recommended results of these methods achieve acceptable accuracy. The most important goal in an RS is to increase the accuracy of items recommended to users ( Vuong Nguyen et al, 2021 ). This means the RS must be designed to collect as much information as possible from users.…”