User Rating Data Research Articles

Recommendation in a Changing World

Users’ preferences, and consequently their ratings and reviews to items, change over time. Likewise, characteristics of items are also time-varying. By dividing data into time periods, temporal Recommender Systems (RSs) improve recommendation accuracy by exploring the temporal dynamics in user rating data. However, temporal RSs have to cope with rating sparsity in each time period. Meanwhile, reviews generated by users contain rich information about their preferences, which can be exploited to address rating sparsity and further improve the performance of temporal RSs. In this article, we develop a temporal rating model with topics that jointly mines the temporal dynamics of both user-item ratings and reviews. Studying temporal drifts in reviews helps us understand item rating evolutions and user interest changes over time. Our model also automatically splits the review text in each time period into interim words and intrinsic words. By linking interim words and intrinsic words to short-term and long-term item features, respectively, we jointly mine the temporal changes in user and item latent features together with the associated review text in a single learning stage. Through experiments on 28 real-world datasets collected from Amazon , we show that the rating prediction accuracy of our model significantly outperforms the existing state-of-art RS models. And our model can automatically identify representative interim words in each time period as well as intrinsic words across all time periods. This can be very useful in understanding the time evolution of users’ preferences and items’ characteristics.

Optimal data collection for informative rankings expose well-connected graphs

Given a graph where vertices represent alternatives and arcs represent pairwise comparison data, the statistical ranking problem is to find a potential function, defined on the vertices, such that the gradient of the potential function agrees with the pairwise comparisons. Our goal in this paper is to develop a method for collecting data for which the least squares estimator for the ranking problem has maximal Fisher information. Our approach, based on experimental design, is to view data collection as a bi-level optimization problem where the inner problem is the ranking problem and the outer problem is to identify data which maximizes the informativeness of the ranking. Under certain assumptions, the data collection problem decouples, reducing to a problem of finding multigraphs with large algebraic connectivity. This reduction of the data collection problem to graph-theoretic questions is one of the primary contributions of this work. As an application, we study the Yahoo! Movie user rating data set and demonstrate that the addition of a small number of well-chosen pairwise comparisons can significantly increase the Fisher informativeness of the ranking. As another application, we study the 2011-12 NCAA football schedule and propose schedules with the same number of games which are significantly more informative. Using spectral clustering methods to identify highly-connected communities within the division, we argue that the NCAA could improve its notoriously poor rankings by simply scheduling more out-of-conference games.

The Research of Broadcast Television Community Discovery Technology Based on Double-weight Gaussian Kernel Similarity

The community discovery technology is proposed based on the double-weight Gaussian kernel similarity to meet the need of intellectual precise delivery business of broadcast television. The text proposes a community technique based on the double-weight Gaussian kernel similarity, according to the user rating data and program broadcast data. It takes user preferences of program category as nodes and the similarity of difierent nodes as edges to build the diagram of TV broadcast community. In order to realize the segmentation of TV users, it adopts the spectral clustering algorithm and transfer the diagram into the matrix of user preferences of program category. In addition, it proposes Gaussian kernel similarity algorithm based on double-weight. For the multivariate data, such as Wine data and Heart data, its veracity can be enhanced by 2% according to the similarity based on the attribute character and the spatial character of the data. Checking the community discovery technique based on double-weight Gaussian kernel similarity, the result shows that it can flnd out user communities of difierent preferences of program category, converge to the global optimal situation and simplify the computational complexity.

Robust Analysis on a Privacy Preserving Recommendation Algorithm under the KNN Attack

— Among algorithms in recommendation system, Collaborative Filtering (CF) is a popular one. However, the CF methods can’t guarantee the safety of the user rating data which cause private preserving issue. In general, there are four kinds of methods to solve private preserving: Perturbation, randomization, swapping and encryption. In this paper, we mimic algorithms which attack the privacy-preserving methods with randomized perturbation techniques. After leaking part of rating history of a customer, we can infer this customer’s other rating history. At the end, we propose an algorithm to enhance the system so as to avoid being attacked.

An Improved Personalized Recommendation System Research

In order to find information of interest and found valuable information resources in enrich Internet data. This paper describes a personalized recommendation system, personalized recommendation system is an intelligent recommendation system to help e-commerce site for customers to provide complete personalized shopping decision support and information services. for the User Rating data extreme sparseness, This paper presents nearest neighbor collaborative filtering algorithm based on project score predicted ,experiments show that this method can improve the quality of recommendation system.

A Personalization Recommendation Algorithm for E-Commerce

Many recommendation systems employ the collaborative filtering technology, which has been proved to be one of the most successful techniques in recommendation systems in recent years, the difficulties of the extreme sparsity of user rating data have become more and more severe. To solve the problems of scalability and sparsity in the collaborative filtering, this paper proposed a personalization recommendation algorithm based on rough set which is proposed, The algorithm refine the user ratings data with dimensionality reduction, then uses a new similarity measure to find the target users’ neighbors, then generates recommendations. To prove our algorithm’s effectiveness, the authors conduct experiments on the public dataset. Theoretical analysis and experimental results show that this method is efficient and effective.

MULTI-GRANULARITY KNOWLEDGE MINING ON THE WEB

We tackle the problem of knowledge mining on the Web. In this paper, we propose MGKM algebraic system for iterative search documents sets, and then develop an approach to extract topics on the web with Multi-Granularity Knowledge Mining algorithm (MGKM). The proposed approach maps the data space of the original method to a vector space of sentence, improving the original DBCO algorithm. We outline the interface between our scheme and the current data Web, and show that, in contrast to the existing approaches, no exponential blowup is produced by the MGKM. Based on the experiments with real-world data sets of 310 users in three study sites, we demonstrate that knowledge mining in the proposed approach is efficient, especially for large-scale web learning resources. According to the user ratings data of four learning sites in the 150 days, the average rate of increase of user rating after the system is used reaches 25.18%.

A Recommendation System Framework Based on Web Mining

Presently recommendation systems have gradually become an important part in E-Commerce, more and more research papers about recommendation systems in E-Commerce appeared in many kinds of conferences and journals. With expanding of E-Commerce it also faces series of challenges. Traditional collaborative filtering recommendation technique is hard to provide recommendation service for unregistered users. To overcome this problem, we suggested a framework of recommendation system based on web mining. It is made up of two parts, offline and online. This method first clustered web usage data, web content data and web structure data respectively, then provided high-quality recommendation services based on mining results. Compared with traditional collaborative filtering techniques, recommendation systems based on web mining are convenient for users because user need not to provide user-rating data explicitly. In end of this paper, accuracy of recommendation system based on web mining was tested and compared with traditional collaborative filtering recommendation system. Testing results showed that, quality of recommendation system based on web mining is better than quality of traditional collaborative filtering recommendation system.

A collaborative filtering recommendation algorithm based on item and cloud model

Recommender system is an important content in the research of E-commerce technology. Collaborative filtering recommendation algorithm has already been used successfully at recommender system. However, with the development of E-commerce, the difficulties of the extreme sparsity of user rating data have become more and more severe. Based on the traditional similarity measuring methods, we introduce the cloud model and combine it with the item-based collaborative filtering recommendation algorithms. The new collaborative filtering recommendation algorithm based on item and cloud model (IC-Based CF) computes the similarity degree between items by comparing the statistical characteristic of items. The experimental results show that this method can improve the performance of the present item-based collaborative filtering algorithm with extreme sparsity of data.

A New Intelligent Topic Extraction Model on Web

We tackle the problem of topic extraction on Web. In this paper, we propose an approach to implementing ontology-based data access in WordNet with the distinguishing feature of optimizing density-based clustering OPTICS algorithm (DBCO) to extract topics. Our solution has the following two desirable properties: i) it uses WordNet for word sense disambiguation of words in the learning resources documents and ii) it mapping the data space of the original method to a vector space of sentence, improving the original OPTICS algorithm. We outline the interface between our scheme and the current data Web, and show that, in contrast to the existing approaches, no exponential blowup is produced by the DBCO. Based on the experiments with a number of real-world data sets of 310 users in three study sites, we demonstrate that topic extraction in the proposed approach is efficient, especially for large-scale web learning resources. According to the user ratings data of four learning sites in the 150 days, the average rate of increase of user rating after the system is used reaches 25.18%.

An Enhanced Similarity Measure Used in Personalized Recommendation Algorithms

Personalized recommendation systems combine the data mining technology with users browse profile and provide recommendation set to user forecasted by their interests. Collaborative filtering algorithm is one of the most successful methods for building personalized recommendation system, and is extensively used in many fields to date. With the development of E-commerce, the magnitudes of users and items grow rapidly, resulting in the extreme sparsity of user rating data. Traditional similarity measure methods work poor in this situation, make the quality of recommendation system decreased dramatically. To alleviate the problem, an enhanced Pearson correlation similarity measure method is introduced in the personalized collaborative filtering recommendation algorithm. The approach considers the common correlation rating of users. The recommendation using the enhanced similarity measure can improve the neighbors influence in the course of recommendation and enhance the accuracy and the quality of recommendation systems effectively.

A Collaborative Filtering Recommendation Algorithm Based on Cloud Model

Recommendation system is one of the most important technologies applied in e-commerce. Similarity measuring method is fundamental to collaborative filtering algorithm,and traditional methods are inefficient especially when the user rating data are extremely sparse. Based on the outstanding characteristics of Cloud Model on the process of transforming a qualitative concept to a set of quantitative numerical values,a novel similarity measuring method,namely the likeness comparing method based on cloud model (LICM) is proposed in this paper. LICM compares the similarity of two users on knowledge level,which can overcome the drawback of attributes’ strictly matching. This work analysis traditional methods throughly and puts forward a novel collaborative filtering algorithm,which is based on the LICM method. Experiments on typical data set show the excellent performance of the present collaborative filtering algorithm based on LICM,even with extremely sparsity of data.