User profile correlation-based similarity (UPCSim) algorithm in movie recommendation system

Abstract

Collaborative filtering is one of the most widely used recommendation system approaches. One issue in collaborative filtering is how to use a similarity algorithm to increase the accuracy of the recommendation system. Most recently, a similarity algorithm that combines the user rating value and the user behavior value has been proposed. The user behavior value is obtained from the user score probability in assessing the genre data. The problem with the algorithm is it only considers genre data for capturing user behavior value. Therefore, this study proposes a new similarity algorithm – so-called User Profile Correlation-based Similarity (UPCSim) – that examines the genre data and the user profile data, namely age, gender, occupation, and location. All the user profile data are used to find the weights of the similarities of user rating value and user behavior value. The weights of both similarities are obtained by calculating the correlation coefficients between the user profile data and the user rating or behavior values. An experiment shows that the UPCSim algorithm outperforms the previous algorithm on recommendation accuracy, reducing MAE by 1.64% and RMSE by 1.4%.

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

“…We select k = 5 from these values to divide the dataset because k = 5 consumes less time than k = 10. In addition, many previous studies [1,3,9,10,12,16] in recommender systems split the dataset into 80%:20% as the training data and testing data.…”

“…Recently, the research conducted by [16] presents a novel similarity metric known as User Profile Correlation-based Similarity (UPCSim). UPCSim improved UPCF, replacing the similarity weight in UPCF (threshold value) with the correlation coefficient between user profile factors and user rating/behavior score.…”

“…The memory-based method consisted of two processes: similarity calculation and rating prediction. In this study, the similarity calculation applies UPCSim [16], which refers to Equation (5).…”

“…S b (x, y) states the user behavior score-based similarity between users x and y, whose formula refers to the Equation (7). Finally, α and β are correlation coefficients between user profile attributes and user rating/behavior scores, calculated using multiple linear regression [16].…”

“…Recently, the development of similarity metrics has considered the user rating score and user behavior score. The similarity metrics that have adopted user behavior scores in calculating similarity are User score Probability Collaborative Filtering (UPCF) [15] and User Profile Correlation-based Similarity (UPCSim) [16]. However, adding the user behavior score variable in the similarity calculation causes the computation to be more complex.…”