Collaborative Filtering Techniques Research Articles

Alleviating sparsity by specificity‐aware ontology‐based clustering for improving web service recommendation

The rapid development of the Internet in recent years has led to a vast increase in the numbers of Web services, which challenge users' capability to find their favorite services quickly and accurately. There is thus an urgent demand for service recommendations that help discover applicable services. Although the collaborative filtering technique is one of the most successful recommendation system technologies, it suffers from data sparsity and cold‐start problems, which in turn lead to inaccurate results. In this article, we deal with these issues by applying a novel ontology‐based clustering approach that uses domain specificity and service similarity for ontology generation. This clustering approach can easily and effectively increase the data density of the user‐service dataset by assuming blank user preferences according to the history of user‐favored domain(s). Then, user ratings are predicted by calculating the trust value between users. The experimental results indicate that the proposed approach can effectively alleviate the data sparsity and cold‐start problems with lower prediction error with the best recommendation performance, and our method performed better than baseline approaches in terms of accuracy and novelty. © 2019 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.

Wireless channel estimation in OFDM systems based on collaborative filtering techniques

Abstract In this paper, a new channel estimation algorithm in Orthogonal Frequency Division Multiplexing (OFDM) systems is proposed. The proposed algorithm is suitable for cases with low density of pilot sub-carriers, where standard interpolation methods (linear, second order and cubic spline interpolation) are inaccurate. The algorithm improves the interpolation methods by employing memory based collaborative filtering (CF) techniques which are less sensitive to the number and location of the pilot subcarriers. CF algorithms are usually used in the context of recommender systems (e-commerce) for predictions of the unknown user-item ratings based on known values of similar users. The advantage of CF is the ability to e ciently produce quality predictions with highly sparse data. Computer simulations are used to verify the proposed channel estimation algorithm and demonstrate that the proposed algorithm improves predictive accuracy metrics, such as Root Mean Squared Error (RMSE), compared to usual estimation methods.

Survey on Recommendation Engines built using Collaborative Filtering Techniques

Survey on Recommendation Engines built using Collaborative Filtering Techniques

Developing a Healthcare Recommender System Using an Enhanced Symptoms-Based Collaborative Filtering Technique

Developing a Healthcare Recommender System Using an Enhanced Symptoms-Based Collaborative Filtering Technique

Variation-Aware Cloud Service Selection via Collaborative QoS Prediction

As the number of cloud services (CSs) offering similar functionality is growing, more attention has been payed on the quality of service (QoS) of CSs. However, in a dynamic cloud environment, the explicit and inherent variation of QoS causes the single CS selection via collaborative filtering techniques (CSS-CFT) to be challenging. A variation-aware approach via collaborative QoS prediction is proposed to select an optimal CS according to users’ non-functional requirements. Based on time series QoS data, this approach utilizes a set of specific cloud models to quantify the variation characteristics of QoS from the four aspects including central tendency, variation range, frequency of variation and period. To exactly identify the neighboring users for a current user, this paper employs the double Mahalanobis distances to measure the similarity of QoS cloud models. The variation-aware CSS-CFT is formulated as a multi-criteria decision-making problem, and an improved TOPSIS method is exploited to solve it, by considering both the objective QoS variation and subjective user preferences during different time periods. The experiments based on a real-world dataset demonstrate that the proposed approach can enhance the accuracy of CSS-CFT in a high-variance environment without noticeable increase of selection time, in comparison to the existing approaches.

Recommending untapped M&A opportunities: A combined approach using principal component analysis and collaborative filtering

This study proposes an analysis framework for recommending untapped mergers and acquisitions (M&A) opportunities by combining principal component analysis (PCA) and collaborative filtering. The framework is particularly suitable for suggesting M&A targets that may possess the complementary knowledge resources for improving performance gains. In this study, PCA was firstly applied to reduce the dimensionality of original data and to identify key knowledge categories of firms. Based on these results, the potential knowledge requirements and priority areas of firm's knowledge base can be located. Collaborative filtering technique was then adopted to a firm's knowledge portfolio, which is represented as a set of SIC (standard industrial classification) codes, for making recommendations. The applicability of the framework was demonstrated using a case study approach based on the analysis of the biotechnology M&A market. The proposed framework would be a good complement to existing practice of identifying novel M&A opportunities and can create a basis for constructing a data-driven business intelligence system. From an expert and intelligent systems point of view, the study provides decision makers with a general and flexible modeling approach for assisting them in the formulation of M&A strategy. It enlarges the possibility of decomposing the problem into manageable units by automating the emulation of human decision making.

DPM-IEDA: Dual Probabilistic Model Assisted Interactive Estimation of Distribution Algorithm for Personalized Search

An interactive evolutionary algorithm (IEA) is powerful for solving personalized search when the user's preference can be well caught, expressed, and applied in the process of searching. Hybrid recommendation by articulating the content-based and collaborative filtering techniques is popular and effective for the personalized recommendation, but has not been developed to improve the performance of IEA for fulfilling the personalized search. Accordingly, we here propose an enhanced interactive estimation of distribution algorithm by designing dual-probabilistic models based on the hybrid recommendation for personalized search. The concept of hybrid personalized search is first defined from the viewpoint of using not only the historical search information but also the social or group preference. A dual-probabilistic model by sufficiently combining the content-based and collaborative filtering is presented and used to design the effective interactive estimation of distribution algorithm (IEDA). The probabilistic model is directly combined with the initialization of IEDA for illuminating the sparsity of the traditional IEA in encoding. The effectiveness of the proposed algorithm in fast and efficient searching with a lower computational cost is experimentally illustrated by two typical personalized searches on movies and TV series described with documents.

Social Information Filtering-Based Electricity Retail Plan Recommender System for Smart Grid End Users

Rapid growth of data in smart grids provides great potentials for the utility to discover knowledge of demand side and design proper demand side management schemes to optimize the grid operation. The overloaded data also impose challenges on the data analytics and decision making. This paper introduces the service computing technique into the smart grid, and proposes a personalized electricity retail plan recommender system for residential users. The proposed personalized recommender system (PRS) is based on the collaborative filtering technique. The energy consumption data of users are firstly collected from the smart meter, and then key energy consumption features of the users are extracted and stored into a user knowledge database (UKD), together with the information of their chosen electricity retail plans. For a target user, the recommender system analyzes his/her energy consumption pattern, find users having similar energy consumption patterns with him/her from the UKD, and then recommend most suitable pricing plan to the target user. Experiments are conducted based on actual smart meter data and retail plan data to verify the effectiveness of the proposed PRS.

Location based Personalized Recommendation systems for the Tourists in India

This study examines the collaborative filtering in recommender system by categorising users according to their choices of place, food, local item purchase, etc. The proposed system will store the opinions of the local users about the sites, foods and products for purchase available in those sites. It uses collaborative filtering technique to find the similar users to a given querying user. The system recommends the best sites along with good foods and products available on those sites according to the recent data. Two hundred (male = 110, female = 90) married individuals from Bhubaneswar, Odisha (India) participated in this survey. Cosine similarity is used in the proposed system to find the similar users of a given input query user. The results revealed that collaborative filtering is the more reliable technique for personalised recommender systems. Experimental results show performance of the proposed system in terms of precision, recall and F-measure values.

FS-CARS: fast and scalable context-aware news recommender system using tensor factorisation

Matrix factorisation is a widely adopted approach of collaborative filtering technique which factorises user-item rating matrix to generate recommendations. User-item rating matrix can be extended to incorporate user's context, resulting in rating tensor which can be factorised to generate better quality context-aware recommendations. Tensor factorisation is computationally intensive task; computational time can be significantly reduced using a distributed and scalable framework. This paper proposes a context-aware news recommender system which classifies news items into different categories and incorporates user's context resulting in rating tensor which is then factorised to generate recommendations. The news items are highly dynamic and are generated in large numbers which can further increase the computational time many fold. To fix the computation time of the process, the proposed system is implemented on distributed and scalable framework of Apache Spark using MLlib library. The proposed recommender system is evaluated for performance and computational time.

Fuzzy Side Information Clustering-Based Framework for Effective Recommendations

Collaborative filtering (CF) is the most successful and widely implemented algorithm in the area of recommender systems (RSs). It generates recommendations using a set of user-product ratings by matching similarity between the profiles of different users. Computing similarity among user profiles efficiently in case of sparse data is the most crucial component of the CF technique. Data sparsity and accuracy are the two major issues associated with the classical CF approach. In this paper, we try to solve these issues using a novel approach based on the side information (user-product background content) and the Mahalanobis distance measure. The side information has been incorporated into RSs to further improve their performance, especially in the case of data sparsity. However, incorporation of side information into traditional two-dimensional recommender systems would increase the dimensionality and complexity of the system. Therefore, to alleviate the problem of dimensionality, we cluster users based on their side information using k-means clustering algorithm and each user's similarity is computed using the Mahalanobis distance method. Additionally, we use fuzzy sets to represent the side information more efficiently. Results of the experimentation with two benchmark datasets show that our framework improves the recommendations quality and predictive accuracy of both traditional and clustering-based collaborative recommendations.

General Collaborative Filtering for Web Service QoS Prediction

To avoid the expensive and time-consuming evaluation, collaborative filtering (CF) methods have been widely studied for web service QoS prediction in recent years. Among the various CF techniques, matrix factorization is the most popular one. Much effort has been devoted to improving matrix factorization collaborative filtering. The key idea of matrix factorization is that it assumes the rating matrix is low rank and projects users and services into a shared low-dimensional latent space, making a prediction by using the dot product of a user latent vector and a service latent vector. Unfortunately, unlike the recommender systems, QoS usually takes continuous values with very wide range, and the low rank assumption might incur high bias. Furthermore, when the QoS matrix is extremely sparse, the low rank assumption also incurs high variance. To reduce the bias, we must use more complex assumptions. To reduce the variance, we can adopt complex regularization techniques. In this paper, we proposed a neural network based framework, named GCF (general collaborative filtering), with the dropout regularization, to model the user-service interactions. We conduct our experiments on a large real-world dataset, the QoS values of which are obtained from 339 users on 5825 web services. The comprehensive experimental studies show that our approach offers higher prediction accuracy than the traditional collaborative filtering approaches.

Collaborative Filtering: Data Sparsity Challenges

Today internet is a place where the huge amount of data is stored, there is need to sift, which create a problem for the internet user, so recommend system solve the problem. A recommendation system is a system that helps a user found the products and content by forecast the user’s rating of each item and showing them the items that they would rate highly. Recommendation systems are everywhere. With online shopping, customer has nearly infinite choices. No one has enough time to try every product for sale. Recommendation systems play an important role to solve the users search the products and content they care about. Recommendation system is a process of filtering the information that deal with information overloaded problems. Recommendation system is important for both user and service provider. It reduces the cost of transaction and selecting item in an online scenario it also improve the quality of decision making process. It is now an effective means for selling their product. So over emphasized of user is not good for recommendation system. To solve the problems of recommendation system like data sparsity we use one of best technique that is collaborative filtering technique.

Automatic cloud resource management for interactive remote geovisualization

AbstractRemote geovisualization has gained momentum to support large‐scale geospatial data analysis and complex decision‐making over the last few years. Cloud computing, due to its capabilities to deliver on‐demand computing resources, has been embraced to develop and deploy interactive and scalable remote geovisualization applications. However, current cloud computing frameworks do not offer a versatile resource management scheme that is readily applicable for online remote visualization services, which usually require maintaining a satisfactory service level over time under dynamic workloads. To address this gap, we propose an automatic cloud resource management approach based on a bi‐level scheduling and horizontal scaling scheme to exploit cloud resources efficiently. At the lower level, a dynamic task‐scheduling scheme using collaborative filtering techniques is proposed to allocate virtual cloud resources to execute sub‐tasks. The scheduling scheme considers spatio‐temporal patterns presented in visualization views. At the upper level, reinforcement learning is adopted to perform resource auto‐scaling based on a reward function that integrates three different facets, namely: time cost, resource cost, and service stability. The original reinforcement learning algorithm is improved in two main aspects: (1) considering the delay of resource provisioning that is common in cloud environments; and (2) using online Gaussian estimation to estimate Q values. Task scheduling and auto‐scaling interact with each other and are integrated to deliver a comprehensive and responsive resource management solution. Experimental results demonstrate that our approach outperforms several existing cloud resource management methods. The proposed approach is also applicable for other interactive visualization applications, which have similar workload characteristics and performance requirements as interactive remote geovisualization.

Improving Collaborative Filtering-Based Image Recommendation through Use of Eye Gaze Tracking

Due to the overwhelming variety of products and services currently available on electronic commerce sites, the consumer finds it difficult to encounter products of preference. It is common that product preference be influenced by the visual appearance of the image associated with the product. In this context, Recommendation Systems for products that are associated with Images (IRS) become vitally important in aiding consumers to find those products considered as pleasing or useful. In general, these IRS use the Collaborative Filtering technique that is based on the behaviour passed on by users. One of the principal challenges found with this technique is the need for the user to supply information concerning their preference. Therefore, methods for obtaining implicit information are desirable. In this work, the author proposes an investigation to discover to which extent information concerning user visual attention can aid in producing a more precise IRS. This work proposes therefore a new approach, which combines the preferences passed on from the user, by means of ratings and visual attention data. The experimental results show that our approach exceeds that of the state of the art.

CFSH: Factorizing sequential and historical purchase data for basket recommendation.

To predict what products customers will buy in next transaction is an important task. Existing work in next-basket prediction can be summarized into two paradigms. One is the item-centric paradigm, where sequential patterns are mined from customers’ transactional data and leveraged for prediction. However, these approaches usually suffer from the data sparseness problem. The other is the user-centric paradigm, where collaborative filtering techniques have been applied on customers’ historical data. However, these methods ignore the sequential behaviors of customers which is often crucial for next-basket prediction. In this paper, we introduce a hybrid method, namely the Co-Factorization model over Sequential and Historical purchase data (CFSH for short) for next-basket recommendation. Compared with existing methods, our approach conveys the following merits: 1) By mining global sequential patterns, we can avoid the sparseness problem in traditional item-centric methods; 2) By factorizing product-product and customer-product matrices simultaneously, we can fully exploit both sequential and historical behaviors to learn customer and product representations better; 3) By using a hybrid recommendation method, we can achieve better performance in next-basket prediction. Experimental results on three real-world purchase datasets demonstrated the effectiveness of our approach as compared with the state-of-the-art methods.

Implementation of Recommender System for Web pages

In Web based applications, Recommender systems have become fundamental information access system. They effectively prune large information spaces which provide appropriate decision making and suggestions, so that the users are directed towards those web pages that best meet their needs, preferences and interests. In web-based context, this can be achieved by basic rough set model and collaborative filtering techniques in decision making of the web pages. The recommender system can be implemented based on two types of techniques which are content based filtering and collaborative filtering. Content based filtering constructs a recommendation on basis of user’s behavior (historical browsing information) and collaborative filtering uses group knowledge to form a recommendation based on like users. Sparsity, the cold-start problem, fraud are main challenges in recommender system.This paper proposes recommendations to the user that varies from one user to another user which is based on the user's profile. (For example, for a search with same keyword a student will be suggested differently compared to a scholar or a teacher).The abstract should state the purpose, approach, results and conclusions of the work. The author should assume that the reader has some knowledge of the subject but has not read the paper. Thus, the abstract should be intelligible and complete in it-self (no numerical references); it should not cite figures, tables, or sections of the paper. The abstract should be written using third person instead of first person.

THE RECOMMENDATION ALGORITHM FOR AN ONLINE ART GALLERY

The paper discusses the need for recommendations and the basic recommendation systems and algorithms. In the second part the design and implementation of the recommender system for online art gallery (photos, drawings, and paintings) is presented. The designed customized recommendation algorithm is based on collaborative filtering technique using the similarity between objects, improved by information from user profile. At the end conclusions of performed algorithm are formulated.

Analysis of similarity measures in user-item subgroup based collaborative filtering via genetic algorithm

Recommender systems based on collaborative filtering (CF) analyze the mutual interests of similar users to predict the ratings of items for the active user (the user whom the prediction is for). This is done in two ways: (1) finding similarity between all users who share the same rating patterns with the active user (2) finding similarity between all pairs of items of different users by building an item-item matrix and thereafter inferring the tastes of the active user. We believe that considering the entire set of items is irrelevant in the prediction process and propose a user-item subgroup based local least squares CF technique that considers only a subset of highly correlated items based on set of similar users. An evolutionary algorithm framework is used to discover the subset of highly correlated items and a local least squares method is used to impute the missing ratings by analyzing the highly correlated user-item subgroup. As far as the experimental setup is concerned, three different similarity measures are used in the implementation of our proposed algorithm and a comparison is made with the traditional local least squares approach as well as with a state-of-the-art CF algorithm. Benchmark datasets like MovieLens are used for the experimental results.

Black-start decision making based on collaborative filtering for power system restoration

Black-start decision making plays an important role in power system restoration after a large-area blackout. The entropy method is one of the most popular methods for determining the weights of indexes in black-start decision making. In spite of its success, the entropy method is affected significantly by normalization methods, which makes the final evaluation results of black-start schemes not reliable. To address this issue, in this article, we propose a novel black-start decision making method based on collaborative filtering. In the proposed method, the values in the decision matrix are withheld, and the collaborative filtering technique is adopted to predict the withheld values. Based on the prediction values and true values, the MAE weights of all the indexes are obtained. Finally, the MAE weights are used to compute the overall assessment value of each black-start scheme. Based on the data of Guangdong power system of China, experiments were carried out to evaluate the performance of the proposed method. Experimental results show that the proposed method is more stable and superior to the existing methods.