|
1.INTRODUCTIONIn recent years, with the rapid development of the Internet, more and more data are stored in the network. The rapid development of web applications, especially mobile applications, enables people to easily browse a large number of online information resources. Therefore, how to recommend resources (e.g., commodities, movies, books, etc.) that meet users’ needs from the vast amount of information resources has become one of the current concerns of researchers1. Recommendation systems play a very effective role in filtering and screening information, and can effectively alleviate the information overload problem. Among the current research on recommendation, collaborative filtering is a basic recommendation algorithm and has been widely used in academia and industry. Despite this, there are still inevitable problems such as data sparsity, cold start, poor real-time performance, and outdated information2. To address the outdated information problem, the traditional collaborative filtering algorithm makes recommendation based on users’ historical rating data and does not fully consider the influence of the time factor3. In this paper, we take the time effect of human interest as the starting point, and construct a time decay function by comparing different time decay factors to combine with collaborative filtering for more time-efficient recommendation. 2.RELATED WORK2.1Collaborative filteringCollaborative filtering is one successful personalized recommendation technology, and is widely used in many fields4. The recommendation process that collaborates with everyone’s feedback, reviews and opinions to filter the massive amount of information together and filter out the information that may be of interest to the target user5.
We use UserCF as the basis for our hybrid algorithm. 2.2User similarity calculationSimilarity calculation is a key step that affects the performance of the recommendation algorithm. There are many methods to calculate similarity, among which the cosine similarity method is the most classical and especially suitable for data sparse situations, so we choose the cosine similarity algorithm to calculate the user similarity6. For users и and v the cosine similarity is calculated as shown in equation (1): where N(u) denotes the set of items which user и has rated, N(v) denotes the set of items which user v has rated, and wuv denotes the cosine similarity between user и and user v. The existing collaborative filtering-based recommendation algorithms have a problem: they always only focus on the similarity between users or resources, but ignore the dynamic changes of users’ interests. But in real life, users’ needs for resources change over time7, and traditional collaborative filtering algorithms only use the user-resource access matrix for recommendation calculation, without considering the specific time of users’ access to resources. When the user’s interest changes, the existing recommendation system cannot detect it in time, so the recommended resources may be largely deviated from the user’s demand. In this paper, we mix time factor and user collaborative filtering (Time-UserCF) to perform analytical comparison experiments. 2.3Time effectTime effect can be seen everywhere in our daily life. For example, dressing styles vary with age; dressing changes with the seasons, and so do fruits. 2.4Preference migrationDue to users’ own reasons, their interests change over time. The user’s preferences directly affect the recommended result set. Here we select the trend of the average rating score change of movies corresponding to user with uid=1086360 from the Netflix dataset between September 2002 and December 2005 as shown in Figure 1. From Figure 1, we can see that the trend of user’s rating score change is different at each time point. Therefore, the corresponding time function should be introduced to simulate the change of user’s interest when rating the items. 3.COLLABORATIVE FILTERING ALGORITHM WITH TIME DECAY FUNCTIONIn this paper, we focus on the hybrid algorithm of UserCF. When using UserCF to recommend items for users, we first find the user set with similar interests to the target user, and then make recommendations for the user based on the behavior of these users8. The formula of user similarity calculation is shown in equation (2): where N(u) denotes the set of items for which user и has generated behavior, N(v) denotes the set of items for which user v has generated behavior, | N(u)∩ N(v)| denotes the number of items for which user и and user v have intersection, and | N(u)∪ N(v)| denotes the number of union items for which user и and user v have generated behavior. When calculating user similarity, a penalty term should be added for popular items, and here the user similarity formula can be modified as equation (3): where N(i) denotes the number of all users who have generated an action on item i. Because the behavior in the most recent period best represents the user’s current interest9, a temporal decay function is added to calculate the similarity between the two users as equation (4): where f(|tui –tvi |) is the time decay function10, which has the form given in equation (5): where α is the time decay factor, tui denotes the time when user и has behavior on item i, and tvi denotes the time when user v has behavior on item i. Since the current rating receives a relatively large influence from the recent ratings of the similar user set. Therefore, the time decay function f(|t0 – tvi |) is also added to calculate the user’s rating of the item, so the final interest degree of user и in item i is obtained in equation (6): where t0 denotes the current time and tvi denotes the time when user v generates an action on item i. 4.EXPERIMENT4.1Experimental environmentUsing a desktop computer with 32GB of RAM and Intel(R) Core(TM) i9-9900K CPU @ 3.60GHz; operating system is windows 10, using PyCharm development tool with python version 3.8. 4.2DatasetDataset is Movie Lens-1M, which is a movie rating dataset collected by the Group Lens research team from the Movie Lens system at different time periods. The data is authentic and valid, including movie attribute information, user demographic information, and movie rating score which is the most critical information for our experiment. 4.3Experimental designThis experiment is modified on the traditional UserCF algorithm. The data are split into training and test sets by calling train_test_split in sklearn, where test_size is 0.1 and random_state is 40. The time decay function f(|tui –tvi|) is added to calculate the user similarity matrix, and by modifying the time decay factor α and the number of nearest neighbor users k for experimental comparison, and we also adding the time decay function f(|t0–tvi|) during recommending. The time decay factor α for training is 0.5 and 5, and the time decay factor β for recommendation is 0.8 and 10, which are divided into two groups, (0.5, 0.8) and (5, 10). These two groups use the number of {5, 15, 25, 35, 45} nearest-neighbor users to compare the experiment with the traditional UserCF. Due to the large number of full-scale test users, only the accuracy, recall and coverage of 10 random users are tested in conducting the evaluation metrics, and then the final evaluation metrics are obtained by repeating the experiment 5 times and calculating the arithmetic mean. 5.RESULTSThe experimental results when the number of nearest neighbor users k = {5, 15, 25, 35, 45} are shown in Figures 2-4. The figure above shows that the accuracy increases slowly with k, and adding the time decay function in the algorithm, the accuracy is improved under the time decay factor (α, β) = (0.5, 0.8), but in the case of time decay factor (α, β) = (5, 10), the accuracy rate has decreased. The figure above shows that the recall increases slowly with k, and adding the time decay function in the algorithm, the recall is improved under the time decay factor (α, β) = (0.5, 0.8), but in the case of time decay factor (α, β) = (5, 10), the recall rate has decreased. The figure above shows that the coverage decreases slowly with k, and adding the time decay function in the algorithm, the coverage is improved to some extent. The time decay factor has little effect on coverage. 6.CONCLUSION AND FUTUREThe experiments show that by modifying the time decay factor and adding the time decay function in the algorithm, the accuracy and recall are improved under the first time decay factor, and the recommendation efficiency has also been improved. However, in the case of the second time decay factor, the accuracy and recall rate are lower than the traditional algorithm, the coverage is both improved compared with the traditional algorithms, and these two time-decay factors have little effect on the coverage. In the future, we will use deep learning methods to improve the collaborative filtering algorithm on temporal context, considering the change of user’s interest to improve the accuracy and coverage, and enhance the system efficiency. REFERENCEYu, M., He, W. T., Zhou, X. C., Cui, M. T., Wu, K. Q. and Zhou, W. J.,
“Overview of recommendation system,”
Computer Application, 1
–16
(2021). Google Scholar
Lan, Y. and Cao, F. F.,
“Research on time weighted collaborative filtering algorithm for film recommendation,”
Computer Science, 44
(04), 295-301
–322
(2017). Google Scholar
Zhang, Z. F., Tang, M. and Li, Q. X.,
“Research on personalized film recommendation algorithm based on time context,”
Journal of Hebei Software Vocational and Technical College, 22
(04), 5
–8
(2020). Google Scholar
Fang, P.,
“An electronic commerce collaborative filtering recommendation algorithm based on user context,”
Applied Mechanics and Materials, 3634 687
–691
(2014). Google Scholar
Xing, C. X., Gao, F. R., Zhan, S. N. and Zhou, Z. Z.,
“Collaborative filtering recommendation algorithm to adapt to the change of user interest,”
Computer Research and Development,
(02), 296
–301
(2007). https://doi.org/10.1360/crad20070216 Google Scholar
Liang, S. Y., Peng, X. L., Qin, B., Lin, W. M. and Hu, Z. N.,
“Collaborative filtering book recommendation based on time context optimization,”
Library Forum, 41
(03), 113
–121
(2021). Google Scholar
Zhao, H. Y., Liu, Q. Y., Chen, Q. K. and Cao, J.,
“Matrix decomposition algorithm based on similarity decay with time,”
Small Microcomputer System, 37
(07), 1474
–1478
(2016). Google Scholar
Liu, Y., Wang, Y., Qi, G. T. and Bao, Z. Y.,
“Collaborative filtering top-N recommendation algorithm in time context,”
Computer Technology and Development, 27
(07), 79
–82
(2017). Google Scholar
Deng, Z. P. and Dai, Z. Q.,
“A catering intelligent recommendation model with improved temporal context correlation,”
Software Guide, 13
(11), 152
–153
(2014). Google Scholar
Xia, Z. Q. and Wang, Z. T.,
“Entropy driven collaborative filtering algorithm integrating time context,”
Journal of Chongqing University of Science and Technology (Natural Science Edition), 21
(03), 73
–77
(2019). Google Scholar
|