Search

Representation learning of a user-item interaction network becomes an effective recommendation method. Most of the existing methods regard the interaction network as a static homogeneous network, ignoring the impact of interaction timing and node heterogeneity. In response to this problem, this paper proposed a recommendation method based on dynamic bipartite network representation learning. Firstly, the method constructed a time-series weighted bipartite network, and then respectively mapped user nodes and item nodes to different vector spaces to preserve the heterogeneity of the network, and aggregated the first-order and high-order neighbor information for center nodes with graph convolution. Finally it used a multi-layer perceptron to learn the nonlinear relationship between the two types of node embeddings and performed top-N recommendation. Experimental results on Amazon and Taobao datasets show that the proposed method is significantly superior than the related methods based on static or heterogeneous network representation learning in HR and NDCG indicators. [ABSTRACT FROM AUTHOR]

Network node prediction research currently focuses on the prediction of source nodes and hidden nodes, but lacks research on prediction of new nodes. This paper took the relational network of papers and keywords as the research object, used keyword combination to predict the emergence of new papers, and carried out the prediction research of new nodes . First, this essay projected and weighted the paper-keyword bipartite network into a keyword relational network, and then used the possib ility of keyword combination to predict the emergence of new papers in the future. There are two aspects to consider to calculate this possibility. One is simila rity, which indicates the tendency of keywords to co-occur ; and the other is mutual exclusion, which describes the tendency of k eywords to exclude each othe r. For example, two keywords with a high degree of concord rarely appear at the same time. Collected the papers and keywords information of the journal to con struct the dataset, it verifies the proposed new paper prediction algorithm, and compared with the existing algorithms . The results show that the node prediction algorithm proposed has better prediction effect. [ABSTRACT FROM AUTHOR]

Network node prediction research currently focuses on the prediction of source nodes and hidden nodes, but lacks research on prediction of new nodes. This paper took the relational network of papers and keywords as the research object, used keyword combination to predict the emergence of new papers, and carried out the prediction research of new nodes. First, this essay projected and weighted the paper-keyword bipartite network into a keyword relational network, and then used the possibility of keyword combination to predict the emergence of new papers in the future. There are two aspects to consider to calculate this possibility. One is similarity, which indicates the tendency of keywords to co-occur; and the other is mutual exclusion, which describes the tendency of keywords to exclude each other. For example, two keywords with a high degree of concord rarely appear at the same time. Collected the papers and keywords information of the journal to construct the dataset, it verifies the proposed new paper prediction algorithm, and compared with the existing algorithms . The results show that the node prediction algorithm proposed has better prediction effect. [ABSTRACT FROM AUTHOR]

Neighborhood-based top-N recommendation techniques for implicit users ' inte raction data are ranking models, of which similarity functions are the important ingredients. Traditional similarity functions suffer from two major issues: sparse data and high dimensional data. Sparse data hinder the recommendation system from scoring on smooth neighborhood, high dimensional data causes the curse of dimensionality problem. This paper proposed a representation-based approach named multi object representation learning ( MO) for recommendation, MO was a bipartite node representation learning algorithm, which embedded the different level of network structures and item ordering information into nodes representations to help leverage the recommendation performance. Experimental results on three real data sets of different scales show that the algorithm has higher accuracy and recall than the commonly used recommendation model based on implicit feedback, especially for large-scale data sets, which can effectively alleviate the problems of matrix sparsity and dimensionality disaster, and improve the recommended performance. [ABSTRACT FROM AUTHOR]

Most recommendation systems based on the network structure suffer from lack of diversity, so that the paper proposed a recommendation system of big data based on multi-weight projection of bipartite network. Firstly, it abstracted the basic information of datasets, and applied items-users lists as an input to Levenshtein distance algorithm to compute similarity of each property. Then, it computed the number of common neighbors of the nodes in the bipartite network, the degree of common neighbors of the nodes in the bipartite network and degree of each node in bipartite network, and also computed the triple weights of each side of the bipartite network. Lastly, it adopted the enhanced bipartite projection technique to abstract the potential links of the bipartite network to realize the link prediction based on similarity. It realized the experiments based on both of big dataset and small dataset. The results show that the proposed algorithm outperforms different kinds of recommendation systems in terms of accuracy and coverage of recommendation, at the same time, it outperforms the other recommendation system based on network structure. [ABSTRACT FROM AUTHOR]

The efficiency of traditional user-based collaborative filtering (user-based CF) recommendation algorithm is reduced with data increasing. This paper proposes a recommendation algorithm based on community detection (RACD) in bipartite networks by introducing bipartite network community detection theory into user-based CF recommendation algorithm. Firstly， the user-item rating matrix is mapped into user-item bipartite network. Then， the community information of each user is obtained by using RACD to divide the user-item network. Finally， the items are recommended to the target user according to other users in the same community. Experiments on real-world classic network datasets show that the RACD can effectively improve real-time recommendation efficiency of the recommendation system. [ABSTRACT FROM AUTHOR]

The usual community detection methods are not applicable to bipartite networks due to their special 2 -mode structure. To identifying the community structure of bipartite networks，this paper proposed a novel algorithm based on Kullback-Leibler (KL) divergence between the 2-mode nodes. According to the connecting conditions between user set and object set， the algorithm obtained the link probability distribution on user set of bipartite networks，and developed K L similarity as a metric to evaluate the difference of node link patterns，and then detected the communities in bipartite networks overcoming the limitation of the 2-mode structure on nodes clustering. The experimental results and analysis in compute-generated and real network all show that this algorithm can effectively mine the meaningful community structures in bipartite networks，and improves the performance of community identification in the accuracy and efficiency. [ABSTRACT FROM AUTHOR]

Migrating ERP (enterprise resource planning) to the cloud step by step is an effective way to mitigate risks, but selecting the functions with priority to move is the premise. This paper begins by putting forward the indicators from traditional ERP usage for measuring information intensity based on information intensity theory. Next it uses the bipartite network method to build the user-function model, defines the information absolute intensity and the information associate intensity in the business areas. Accordingly it presents the selection method of migrating ERP to the cloud step by step. Finally, it did an empirical analysis on the real ERP log data from a large scale manufacturing enterprise of shipbuilding industry, and validates the proposed selection method. [ABSTRACT FROM AUTHOR]

To tackle the network congestion issue caused by redundant traffic by users' repeated access to the same network resources, this paper proposed a weighted bipartite network (WBN) model which evolved with time. Through the analysis of users' network behavior and redundant traffic' formation and evolution mechanism, WBN model adopted ways of preferential connection and topology growth to complete network evolution. Simulation experiment results show that strength of user nodes has experienced from the exponential distribution to power-law distribution and then random distribution. It effectively simulates the evolution of real network redundant traffic. Simulation results are also consistent with theoretical analysis and show that WBN model is accurate and practical. [ABSTRACT FROM AUTHOR]

In the age of Internet, the overloaded information turned the information filtering to be a widely discussed and urgent problem. Considering the popularity of objects, this paper presented a non-equilibrium heat conduction information filtering method, in which introduced a tunable parameter λ to control the influence of popularity. The simulation results suggest that,with the optimal value λopt , the precision and recall are improved by 228.2% and 228.4% respectively for the MovieLens system. As to the Amazon system, those are improved by 162.7% and 162.8% respectively. The presented algorithm can improve the performance of recommendation, which suggests the importance of the popularity on users' selecting process. [ABSTRACT FROM AUTHOR]