Search

The primary key in novel recommendation is to judge whether an item is "new" for target users. The novelty of recommended items to users changes with time. Therefore, the effects of time factors on user preference, as well as the novelty and accuracy of the recommended items, have become key to novelty recommendation research. To increase novelty based on recommendation accuracy, a novel recommendation algorithm based on time-aware was proposed. First, a time attenuation parameter was introduced into a traditional algorithm to strengthen the weights of recently adopted items. Next, the calculation of innovation adoption activism was adjusted through time information. Differences among recommended items were calculated by combining the time of items that have been adopted by users and the quantity of relevant categories. Lastly, the novelty of recommended items was calculated by the result fusion strategy for Top-N recommendation. Meanwhile, the effectiveness of the proposed algorithm was verified by conducting an offline experiment. Results demonstrate that the novelty of the recommendation list by the proposed algorithm is significantly higher than that of traditional algorithms. Novelty when the number of items in the candidate set reaches 300-400 reaches the peak. The recall rate of unknown items increases by 25%, whereas the recall rate of known items decreases by 60%. The average popularity of the recommendation list decreases by 50%, and the coverage rate increases by 200%. The proposed algorithm improves the ability of the recommendation system (RS) to explore novel items. This study provides references for improving the satisfaction of users with RSs. [ABSTRACT FROM AUTHOR]

With the rapid development of information technology and Internet computing, recommendation systems of various forms have been used in nearly all large-scale e-commerce platforms to different degrees. Besides accuracy, novelty also influences the satisfaction of users with these recommendation systems. To be considered novel, an item should show some differences from the preferences of users. To recommend diverse items under the precondition of guaranteeing accuracy, a clustering-based novel recommendation algorithm was proposed in this study. User preference was modeled by using global clustering and related-items clustering (RC), weighted distance was employed to calculate the dissimilarity of the to-be-recommended items from user preference, and item novelty was measured by combining satisfactory and unknown, followed by recommendation. The effectiveness of the proposed algorithm was verified by conducting an offline experiment. Results show that the recall rate of unknown items of RC increases by 15% whereas that of known items decreases by 60%, thereby indicating that the novelty of the recommendation result is remarkably improved. Meanwhile, the coverage rate increases by approximately 200%, thereby indicating that the proposed recommendation system demonstrates an improved ability to recommend long-tail items. This study provides references for improving the satisfaction of users with recommendation systems. [ABSTRACT FROM AUTHOR]

Determining whether an item is "novel" for target users is a crucial problem and understanding their preference and awareness degree of the item has been widely investigated in novelty recommendations. A recommendation algorithm based on "diffusion of innovation" theory was proposed in this study to verify the novelty of recommendation results on the precondition of ensuring their accuracy. Items were clustered using the K-means clustering algorithm. Users' positivity of innovation adoption for each item was calculated on the basis of the items adopted by the users, and then the novelty of the items for target users was calculated by combining their popularity change, user preference, and user difference. Results of traditional recommendation algorithms were integrated for recommendation on the basis of fusion strategy results. The effectiveness of the proposed algorithm was verified through an offline experiment. Results indicate that the novelty of the recommendation list of the proposed algorithm is remarkably higher than that of traditional algorithms. The novelty is high when the quantity of alternative sets reaches 400 or more, where the average popularity of the recommendation list declines by 40%, and the coverage is elevated by 100%, thereby improving the ability of the proposed system to extract all kinds of items. This study serves as reference for the improvement of user satisfaction with recommendation systems. [ABSTRACT FROM AUTHOR]