A Multi-attribute Collaborative Filtering Recommendation Algorithm Based on Improved Group Decision-Making

Personalized recommendation system is an important means for people to get interested information and product quickly. This traditional user-based collaborative filtering algorithm cost too much computation on similarity calculation. In order to solve this problem, a new collaborative filtering recommendation algorithm based on K-Means clustering of user’s attributes is proposed. In this algorithm, the longitude and latitude of users’ are first clustered, and then the similarity of users’ are calculated within each cluster. Finally, parallelization of this proposed algorithm on Spark is implemented. Experiments show that the user attributes-based collaborative filtering has satisfied performance.

Currently collaborative filtering is widely used in e-commerce, digital libraries and other areas of personalized recommendation service system. Nearest-neighbor algorithm is the earliest proposed and the main collaborative filtering recommendation algorithm, but the data sparsity and cold-start problems seriously affect the recommendation quality. To solve these problems, A collaborative filtering recommendation algorithm based on users' social relationships is proposed. 0n the basis of traditional filtering recommendation technology, it combines with the interested objects of user's social relationship and takes the advantage of the tags to projects marked by users and their interested objects to improve the methods of recommendation. The experimental results of MAE ((Mean Absolute Error)) verify that this method can get better quality of recommendation.

With the speedy development of network, information technology has provided an unmatched amount of information resources. It has also led to the problem of information overload. However, people experiences and knowledge often do not enough to process the vast amount of usable information. Thus, approaches to help find resources of interest have attracted much attention from researchers. And recommender systems have arrived to solve this problem. Recommender system plays an important role mainly in an electronic commerce environment as a new marketing strategy. Although a varied of recommendation techniques has been developed recently, collaborative filtering has been known to be the most successful recommendation techniques and has been used in a number of different applications. But traditional collaborative filtering recommendation algorithm has the problem of sparsity. Aiming at the problem of data sparsity for personalized filtering systems, a collaborative filtering recommendation algorithm based on user rating similarity and user attribute similarity is given. This approach not only considers the user item rating information, but also takes into account the user attribute.

Collaborative filtering, which makes personalized predictions by learning the historical behaviors of users, is widely used in recommender systems. It makes the prediction and recommend by similarity of users, and it can handle the various work. But the traditional collaborative filtering ignores the connection of users and items. Affect the recommendation's results. To find similar users by measuring the customer relationship between the neighbors can improve the accuracy of prediction user interests'. Then it can improve the accuracy of the recommendation. So collaborative filtering recommendation algorithm based on MDP model is proposed. It can find the connection of users purchase and next purchase. So it can predict users next purchase. Then can recommend items to users. The test results shows the algorithm of this paper have more accuracy.

The traditional collaborative filtering recommendation algorithm can be divided into the user-based and the item-based two methods, which only uses the information in the rating matrix. Because of the limitation of the information capacity they used, it is difficult to further improve the accuracy of the recommendation, and cold start problem also affects the normal operation of the recommendation system. This paper presented a collaborative filtering recommendation algorithm (UISA) fusing user-based, item-based and social networks data. The algorithm uses the data of the neighbor relations in social networks, calculating the users' friends not reflected in the rating matrix. At the same time, we can calculate the similarity between items by using the data of item text in social networks, mining similar items not reflected in the rating matrix. In this way, it can fundamentally expand available information capacity of the traditional filtering collaboration recommendation algorithms, improve the recommendation accuracy, alleviate cold start problem. Experimental results based on KDD CUP 2012 real data show that compared with the traditional collaborative filtering system, this system has obvious advantages in the recommendation accuracy and ease of cold start.

Recommender system is one of the most important technologies in electronic commerce. And the collaborative filtering is almost the popular approach used in the recommender systems. With the development of electronic commerce systems, the magnitudes of users and items grow rapidly, resulted in the extreme sparsity of user rating data set. Traditional similarity measure methods work poor in this situation, make the quality of recommendation system decreased dramatically. Sparsity of users’ ratings is the major reason causing the poor quality. To address this issue, a collaborative filtering recommendation algorithm based on singular value decomposition (SVD) smoothing is presented. This approach predicts item ratings that users have not rated by the employ of SVD technology, and then uses Pearson correlation similarity measurement to find the target users’ neighbors, lastly produces the recommendations. The collaborative filtering recommendation algorithm based on SVD smoothing can alleviate the sparsity problems of the user item rating dataset, and can provide better recommendation than traditional collaborative filtering algorithms.

Collaborative filtering recommendation is a popular recommendation algorithm in electronic commerce, but the disadvantage of cold start still exists in the reason of new coming users and items. CBR (case-based reasoning) is used to get source case in history according to the notation of target case, and the source case plays a guiding role in solving the problem of the target case. It is a useful algorithm to evaluate the solution of target case, and explain the abnormal phenomenon of target case. In this paper we applied case-based reason with forgetting mechanism to solve the cold start problem in collaborative filtering, to deduce the score of items which is not scored by user, and then to recommend items with TOP-N collaborative filtering. Experimental results show that the proposed collaborative filtering combining case-based reasoning could significantly ease cold start problem.

This paper presents a novel collaborative filtering recommendation algorithm based on field authorities which simulates the real life word of mouth recommendation mode. It uses the specialistic knowledge from field authorities of different genres, and successfully addresses data sparsity and noise problems existing in traditional collaborative filtering. Meanwhile it also improves prediction accuracy and saves computational overhead effectively. Experiments on MovieLens datasets show that the accuracy of our algorithm is significantly higher than collaborative filtering approach based on experts, and has larger scope because of no external data limitations. Meanwhile, compared to traditional k-NN collaborative filtering, our algorithm has a better performance both in MAE and precision experiments, and the computational overhead has a decrease of 19.2% while they provide the same accuracy level.

It is meaningful to use data mining technology to predict the type of lawsuit which a company may receive so that enterprises can avoid lawsuit risks. So we propose a corporation lawsuit prediction algorithm based on guiding learning and collaborative filtering recommendation. Firstly, we use the adaptive synthetic sampling approach (ADASYN) to generate more synthetic data for different minority classes according to their different level of difficulty in learning, so that the training would focus on these minority classes that are difficulty to learn and reduce the learning bias introduced by the imbalance of data distribution. Secondly, for the sake of solving the problem that the insufficient samples make it difficult for the model to learn enough knowledge resulting in a large fluctuation of final scores during the training and poor model stability, we use guiding learning to integrate the basic knowledge of all types of lawsuit a company may receive in the future obtained by the multi-label classification model into the training process of TOP-1 and TOP-2 predictive models. Finally, in order to further improve the prediction accuracy, we use the collaborative filtering recommendation algorithm (CFRA) to select the most similar sample with each test sample from the training set, and the lawsuit type of the selected sample is directly used as the predicted lawsuit type of the corresponding test sample, thereby improving the total prediction accuracy. The experimental results show that the proposed algorithm can effectively predict the most probable lawsuit types of the Top2 for corporations.

In the collaborative filtering algorithm, the user similarity and the selection of the nearest neighbor are the two most important links in the whole algorithm. The traditional collaborative filtering algorithm only relies on user ratings to calculate the similarity between users and then find each user’s K neighbors, and finally according to the K neighbor set recommended. However, in the face of large data processing, the traditional collaborative filtering algorithm is sparsely populated, resulting in the recommendation which is not obvious. A collaborative filtering algorithm for hybrid similarity is proposed for this problem. The algorithm is mainly from the user similarity degree of similarity, user attribute similarity and similarity of user purchase records, and many experiments to adjust the weight between them, the use of dynamic selection of the collection method to find the user’s nearest neighbor, so as to users Recommend more suitable items, improve the recommended quality of the algorithm. The experimental results show that the method proposed in this paper has lower MAE value than the traditional co-filtering algorithm, and the recommended quality is improved.

Everyday there are masses of information generated and the existence of a large amount of information makes it hardly to mining the wanted information. Personalized recommendation is the process to alleviative the problem. Collaborative filtering is one of the most popular technologies in the personal recommendation system. As the user rating matrix becoming extremely sparsity, traditional collaborative filtering recommendation algorithm calculates similarity between items using the rating data, and it does not consider the semantic relationship between different items, thus recommendation quality is very poor. To solve this problem, the paper combines the item semantic similarity and the item rating similarity, which takes into account the influence of item semantic and user rating to enhance the item-based collaborative filtering. The personalized collaborative filtering recommendation algorithm combining the item semantic similarity and item rating similarity can mitigate the sparsity problem in the electronic commerce recommender systems.

Aiming at the problem of lack of time information in traditional collaborative filtering recommendation algorithm, the dynamic characteristics of user preference drift over time are analyzed. An improved scheme of collaborative filtering algorithm based on time characteristic is proposed. In calculating the predictive score, taking into account the characteristics of user interest decay over time, the time decay function is introduced as the weight of the loss function, the lower the weight for the interaction farther from the present time, the closer the interaction is with the higher weight, Element-wise Alternating Least Squares (eALS) is used to improve model training speed when updating models. Using the public dataset Yelp to conduct experiments, compared with the traditional RCD method, the predicted hit rate using the eALS method is significantly improved, and the T-eALS method with the added time information makes the hit rate increase significantly, and the experimental results show that this method can effectively integrate time information and provide users with more accurate recommendations.

Collaborative filtering recommendation technology is successfully used in personalized recommendation services. Since the magnitudes of users and commodities in E-commerce system has increased dramatically, the user rating data in the entire item space become extremely sparse. There is a certain deviation while using traditional similarity measure methods, which reduces the recommendation accuracy for the recommendation systems. To overcome the shortages of the traditional similarity measures under such conditions, this paper proposes using similarity impact factor to improve similarity measures in collaborative filtering recommendation algorithms. The experimental results show that the factor can effectively improve the similarity measure result while user rating data are extremely sparse, and significantly improve the accuracy of the recommendation systems.

Collaborative Filtering (CF) is a successful technology that has been implemented in E-commerce recommender systems. However, the risks of shilling attacks have already aroused increasing concerns of the society. Current solutions mainly focus on attack detection methods and robust CF algorithms that have flaws of unassured prediction accuracy. Furthermore, attack detection methods require a threshold to distinguish normal users from fake users and suffer from the problems of false positive if the threshold is too high and false negative if too low. This paper proposes a soft-decision method, Neighbor Selection with Variable-Length Partitions (VLPNS), to reduce false positive rate through marking suspicious fakers instead of deleting them directly such that misclassified normal users can still contribute to the similarity calculation. The method works as follows: First, it gets user's suspicion probability by applying SVM. It then generates partitions of variable sizes from which different numbers of neighbors can be selected by using the bisecting c-means clustering algorithm. Finally, it chooses neighbors considering the user's suspicion degree and similarity with target user at the same time. Theoretical and experimental analysis show that our approach ensures an excellent prediction accuracy against shilling attacks.

This paper fully studied in the Collaborative Filtering Recommendation Algorithm and k-means clustering algorithm, on this basis, Proposed the establishment of recommendation system based on the collaborative Filtering Recommendation Algorithm of k-means clustering algorithm. This can significantly reduce the seeking time of algorithm to find the nearest neighbor time, so as to raise the speed of the E-commerce recommendation system.