Collaborative Filtering Recommender Systems Research Articles

Variational autoencoder Bayesian matrix factorization (VABMF) for collaborative filtering

Probabilistic matrix factorization (PMF) is the most popular method among low-rank matrix approximation approaches that address the sparsity problem in collaborative filtering for recommender systems. PMF depends on the classical maximum a posteriori estimator for estimating model parameters; however, these approaches are vulnerable to overfitting because of the nature of a single point estimation that is pursued by these models. An alternative approach to PMF is a Bayesian PMF model that suggests the Markov chain Monte Carlo algorithm as a full estimation for approximate intractable posterior over model parameters. However, despite its success in increasing prediction, it has a high computational cost. To this end, we proposed a novel Bayesian deep learning-based model treatment, namely, variational autoencoder Bayesian matrix factorization (VABMF). The proposed model uses stochastic gradient variational Bayes to estimate intractable posteriors and expectation–maximization-style estimators to learn model parameters. The model was evaluated on the basis of three MovieLens datasets, namely, Ml-100k, Ml-1M, and Ml-10M. Experimental results showed that our proposed VABMF model significantly outperforms state-of-the-art RS.

Evaluation of Collaborative Filtering for Recommender Systems

Recently, due to the increasing amount of data on the Internet along with the increase in products’ purchasing via e-commerce websites, Recommender Systems (RS) play an important role in guiding customers to buy products they may prefer. Furthermore, these systems help the companies to advertise their products to the most potential customers, and therefore raise their revenues. Collaborative Filtering (CF) is the most popular RS approach. It is classified into memory-based and model-based filtering. Memory-based filtering is in turn classified into user-based and item-based. Several algorithms have been proposed for CF. In this paper, a comparison has been performed between different CF algorithms to assess their performance. Specifically, we evaluated K-Nearest Neighbor (KNN), Slope One, co-clustering and Non-negative Matrix Factorization (NMF) algorithms. KNN algorithm is representative of the memory-based CF approach (both user-based and item-based). The other three algorithms, on the other hand, are under the model-based CF approach. In our experiments, we used a popular MovieLens dataset based on six evaluation metrics. Our results reveal that the KNN algorithm for item-based CF outperformed all other algorithms examined in this paper.

Research Problems in Recommender systems

With continuous growth of web applications around the globe, it is a challenge to find the suitable information needed for the user in a limited time.Number of handheld mobile devices is increasing and most of the business revolves around the correct search of the data. Without a proper recommender system it is very difficult to get required information from the web applications. Web applications use recommender systems to provide suitable data to users based on their choices and interests. For different kinds of needs different types of recommender systems have been proposed. Two most basic types of recommender systems are collaborative filtering recommender system and content based recommender system. Sometimes these two recommender systems are combined to increase the efficiency of a recommender system The generated new recommender system is known as hybrid recommender system.The purpose of this paper is to help readers understand the basics of recommender systems. This paper identifies key areas of research openly available for new researchers. After reading this paper new researchers can understand basic problems of recommender systems which need improvement and hence they can make those problems their area of research.

Intent-Based Network Policy to Solution Architecting Recommendations

The semantic diversity of policies written by people with different IT literacy to achieve certain network security or performance goals created ambiguity to otherwise straightforward solution implementations. In this project, an intent-aware solution recommender is designed to decode semantic cues in network policies written by various demographics for robust solution recommendations. A novel policy analyzer is designed to extract the intrinsic networking intents from ICT policies to provide context-specific solution recommendations. A custom network-aware intent recognizer is trained on a small keywords-to-intents dataset annotated by domain experts using NLP algorithms in AWS comprehend. The bin-of-words model is then used to classify sentences in the policies into predicted ‘intent' class. A collaborative filtering recommendation system using crowd-sourced ground-truth is designed to suggest optimal architecting solutions to achieve the requirements outlined in ICT policies.

Differentially private user-based collaborative filtering recommendation based on [formula omitted]-means clustering

Collaborative filtering (CF) recommendation is well-known for its outstanding recommendation performance, but previous researches showed that it could cause privacy leakage for users due to k-nearest neighboring (KNN) attacks. Recently, the notion of differential privacy (DP) has been applied to privacy preservation in recommendation systems. However, as far as we know, existing differentially private CF recommendation systems degrade the recommendation performance (such as recall and precision) to an unacceptable level. In this paper, to address the performance degradation problem, we propose a differentially private user-based CF recommendation system based on k-means clustering (KDPCF). Specifically, to improve the recommendation performance, KDPCF first clusters the dataset into categories by k-means clustering and appropriately adjusts the size of the target category to which the target user belongs, so that only users in the well-sized target category are used for recommendation. Then, it selects efficiently a set of neighbors from the target category at one time by employing only one instance of exponential mechanism instead of the composition of multiple ones, and then uses a CF algorithm to recommend based on this set of neighbors. We theoretically prove that our system achieves differential privacy. Empirically, we use two public datasets to evaluate our recommendation system. The experimental results demonstrate that our system has a significant performance improvement compared to existing ones.

Deep learning feature selection to unhide demographic recommender systems factors

Extracting demographic features from hidden factors is an innovative concept that provides multiple and relevant applications. The matrix factorization model generates factors which do not incorporate semantic knowledge. This paper provides a deep learning-based method: DeepUnHide, able to extract demographic information from the users and items factors in collaborative filtering recommender systems. The core of the proposed method is the gradient-based localization used in the image processing literature to highlight the representative areas of each classification class. Validation experiments make use of two public datasets and current baselines. Results show the superiority of DeepUnHide to make feature selection and demographic classification, compared to the state of art of feature selection methods. Relevant and direct applications include recommendations explanation, fairness in collaborative filtering and recommendation to groups of users.

A survey of attack detection approaches in collaborative filtering recommender systems

Nowadays, due to the increasing amount of data, the use of recommender systems has increased. Therefore, the quality of the recommendations for the users of these systems is very important. One of the recommender systems models is collaborative filtering (CF) which uses the ratings given by the users to the items. But many of these ratings may be noisy or inaccurate so they reduce the quality of the recommendations. Sometimes users, using fake profiles, try to change the recommendations in their favor. Since satisfaction and trust in such systems are very important and useful, it would be better to find a way to identify these types of users. Despite numerous studies on CF recommender systems, the design of a robust recommender system is still a challenging problem. In this paper, we have analyzed the 25 previous samples of research on collaborative filtering recommender system (CFRS) for attack detection from 2009 to 2019. Most of these papers focus mainly on movie recommendations. According to these analyzes, we have categorized attack detection methods on CFRS in four categories: clustering, classifying, feature extraction and probabilistic approaches. The evaluation measures, the dataset, and attacks features used in the attack detection approaches are discussed.

Job matcher: A web application job placement using collaborative filtering recommender system

Job matcher: A web application job placement using collaborative filtering recommender system

A collaborative filtering recommender system using genetic algorithm

This paper presents a novel genetic-based recommender system (BLIGA) that depends on the semantic information and historical rating data. The main contribution of this research lies in evaluating the possible recommendation lists instead of evaluating items then forming the recommendation list. BLIGA utilizes the genetic algorithm to find the best list of items to the active user. Thus, each individual represents a candidate recommendation list. BLIGA hierarchically evaluates the individuals using three fitness functions. The first function uses semantic information about items to estimates the strength of the semantic similarity between items. The second function estimates the similarity in satisfaction level between users. The third function depends on the predicted ratings to select the best recommendation list.BLIGA results have been compared against recommendation results from alternative collaborative filtering methods. The results demonstrate the superiority of BLIGA and its capability to achieve more accurate predictions than the alternative methods regardless of the number of K-neighbors.

Neural embedding collaborative filtering for recommender systems

The main purpose of collaborative filtering algorithm is to provide a personalized recommender system based on past interactions of each user (e.g., clicks and purchases). Among various collaborative filtering techniques, matrix factorization is widely adopted in diverse applications. This technique has superior characteristics, including applying latent feature vectors to represent users or items and projecting users and items into a shared latent feature space. In the present study, a matrix factorization model with the neural embedding called the neural embedding collaborative filtering (NECF) is proposed. In order to evaluate the performance of the proposed method, a probabilistic auto-encoder is initially applied to achieve unsupervised learning to generate the neural embedding vector from the user–item data. Secondly, these vectors are combined with a regression model based on single point negative sampling to represent the latent feature vectors of the user with regression coefficients. Moreover, an inner product is applied on latent features of users and items to determine the correlations between them. It should be indicated that the NECF is generic so that it can express and generalize the matrix factorization under its framework. In the present study, a ridge regression learning is applied on latent features of each user. The experimental results on two benchmark data sets show that the proposed model outperforms other state-of-the-art methods.

Electricity plan recommender system with electrical instruction-based recovery

Several electricity tariffs have emerged for Demand Side Management (DSM) and residential customers are faced with challenges to choose the plan satisfying their personal needs. Electricity Plan Recommender System (EPRS) can alleviate the problem. This paper proposes a novel EPRS model named EPRS with Electrical Instruction-based Recovery (EPRS-EI), which is a dual-stage model consisting of feature formulation stage and recommender stage. In the feature formulation stage, matrix recovery with electrical instructions is applied to recover appliance usages, and the recovered data is set as features representing customers’ living patterns. In the recommender stage, Collaborative Filtering Recommender System (CFRS) based on K-Nearest Neighbors (KNN) and adjusted similarity is applied to recommend personal electricity plans to customers based on the above features. Different from other EPRS models, EPRS-EI is the first model utilizing matrix recovery methods and similarity computation with electrical instructions. With these electrical instructions, the proposed model is able to utilize more explicit features and recommend more personalized plans. We then apply EPRS-EI to predict the testing customers’ preference for electricity plans. Simulation results on recovering electricity data and their applications in EPRS confirm the effectiveness of the proposed methods in comparison to state-of-the-art methods, with 93.56%–94.85% customers correctly recommended.

Leveraging User Comments for Recommendation in E-Commerce

Collaborative filtering recommender systems traditionally recommend products to users solely based on the given user-item rating matrix. Two main issues, data sparsity and scalability, have long been concerns. In our previous work, an approach was proposed to address the scalability issue by clustering the products using the content of the user-item rating matrix. However, it still suffers from these concerns. In this paper, we improve the approach by employing user comments to address the issues of data sparsity and scalability. Word2Vec is applied to produce item vectors, one item vector for each product, from the comments made by users on their previously bought goods. Through the user-item rating matrix, the user vectors of all the customers are produced. By clustering, products and users are partitioned into item groups and user groups, respectively. Based on these groups, recommendations to a user can be made. Experimental results show that both the inaccuracy caused by a sparse user-item rating matrix and the inefficiency due to an enormous amount of data can be much alleviated.

A collaborative filtering recommendation system with dynamic time decay

The collaborative filtering (CF) technique has been widely utilized in recommendation systems due to the precise prediction of users' interests. Most prior CF methods adapted overall ratings to make predictions by collecting preference information from other users. However, in real applications, people’s preferences usually vary with time; the traditional CF could not properly reveal the change in users’ interests. In this paper, we propose a novel CF-based recommendation, dynamic decay collaborative filtering (DDCF), which captures the preference variations of users and includes the concept of dynamic time decay. We extend the idea of human brain memory to specify the level of a user’s interests (i.e., instantaneous, short-term, or long-term). According to different interest levels, DDCF dynamically tunes the decay function based on users’ behaviors. The experimental results show that DDCF with the integration of the dynamic decay concept performs better than traditional CF. In addition, we conduct experiments on real-world datasets to demonstrate the practicability of the proposed DDCF.

Integrating Community Context Information Into a Reliably Weighted Collaborative Filtering System Using Soft Ratings

In this paper, we aim at developing a new collaborative filtering recommender system using soft ratings, which is capable of dealing with both imperfect information about user preferences and the sparsity problem. On the one hand, Dempster–Shafer theory is employed for handling the imperfect information due to its advantage in providing not only a flexible framework for modeling uncertain, imprecise, and incomplete information, but also powerful operations for fusion of information from multiple sources. On the other hand, in dealing with the sparsity problem, community context information that is extracted from the social network containing all users is used for predicting unprovided ratings. As predicted ratings are not a hundred percent accurate, while the provided ratings are actually evaluated by users, we also develop a new method for calculating user–user similarities, in which provided ratings are considered to be more significant than predicted ones. In the experiments, the developed recommender system is tested on two different data sets; and the experiment results indicate that this system is more effective than CoFiDS, a typical recommender system offering soft ratings.

Financial Inclusion: An Application of Machine Learning in Collaborative Filtering Recommender Systems

In the current scenario in finance, data play a major role for predicting stock market as well as verious financial instruments. For the estimation of financial data, the various algorithms and models have been used. The use of the advising method has been used in this paper. The advising programs are one of the main methodologies used in the present market scenario with machine learning technologies. This paper focuses on the impact of financial inclusion in Odisha using a machine learning approach such as the classification of k-Nearest Neighbors (k-NN). For financial inclusion systems, machine learning has become a commonly used method. The result takes into the ATMs, Banks and BCs ranking in different districts of Odisha. We used the k-Nearest Neighbor's machine learning methodology classification algorithm to characterize the recommendation system based on users of the mentioned populations. Using our approach we equate conventional collective filtering. Our results show that the linear algorithm is more reliable than the current algorithm and is more efficient and stable than current methods.

Diversity Balancing for Two-Stage Collaborative Filtering in Recommender Systems

Conventional recommender systems are designed to achieve high prediction accuracy by recommending items expected to be the most relevant and interesting to users. Therefore, they tend to recommend only the most popular items. Studies agree that diversity of recommendations is as important as accuracy because it improves the customer experience by reducing monotony. However, increasing diversity reduces accuracy. Thus, a recommendation algorithm is needed to recommend less popular items while maintaining acceptable accuracy. This work proposes a two-stage collaborative filtering optimization mechanism that obtains a complete and diversified item list. The first stage of the model incorporates multiple interests to optimize neighbor selection. In addition to using conventional collaborative filtering to predict ratings by exploiting available ratings, the proposed model further considers the social relationships of the user. A novel ranking strategy is then used to rearrange the list of top-N items while maintaining accuracy by (1) rearranging the area controlled by the threshold and by (2) maximizing popularity while maintaining an acceptable reduction in accuracy. An extensive experimental evaluation performed in a real-world dataset confirmed that, for a given loss of accuracy, the proposed model achieves higher diversity compared to conventional approaches.

Resolving data sparsity and cold start problem in collaborative filtering recommender system using Linked Open Data

The web contains a huge volume of data, and it's populating every moment to the point that human beings cannot deal with the vast amount of data manually or via traditional tools. Hence an advanced tool is required to filter such massive data and mine the valuable information. Recommender systems are among the most excellent tools for such a purpose in which collaborative filtering is widely used. Collaborative filtering (CF) has been extensively utilized to offer personalized recommendations in electronic business and social network websites. In that, matrix factorization is an efficient technique; however, it depends on past transactions of the users. Hence, there will be a data sparsity problem. Another issue with the collaborative filtering method is the cold start issue, which is due to the deficient information about new entities. A novel method is proposed to overcome the data sparsity and the cold start problem in CF. For cold start issue, Recommender System with Linked Open Data (RS-LOD) model is designed and for data sparsity problem, Matrix Factorization model with Linked Open Data is developed (MF-LOD). A LOD knowledge base “DBpedia” is used to find enough information about new entities for a cold start issue, and an improvement is made on the matrix factorization model to handle data sparsity. Experiments were done on Netflix and MovieLens datasets show that our proposed techniques are superior to other existing methods, which mean recommendation accuracy is improved.

Data poisoning attacks on neighborhood‐based recommender systems

AbstractNowadays, collaborative filtering recommender systems have been widely deployed in many commercial companies to make profit. Neighborhood‐based collaborative filtering (CF) is common and effective. To date, despite its effectiveness, there has been little effort to explore their robustness and the impact of data poisoning attacks on their performance. Can the neighborhood‐based recommender systems be easily fooled? To this end, we shed light on the robustness of neighborhood‐based recommender systems and propose a novel data poisoning attack framework, encoding the purpose of attack and constraint against them. We first illustrate how to calculate the optimal data poisoning attack, namely, UNAttack. We inject a few well‐designed fake users into the recommender systems such that target items will be recommended to as many normal users as possible. Extensive experiments are conducted on three real‐world datasets to validate the effectiveness and the transferability of our proposed method. In addition, some interesting phenomena can be found. For example, (i) neighborhood‐based recommender systems with Euclidean distance‐based similarity have strong robustness and (ii) the fake users can be transferred to attack the state‐of‐the‐art CF recommender systems such as neural CF and Bayesian personalized ranking matrix factorization.

Cross Attentive Collaborative Filtering for Recommender Systems

The information on the Internet is increasingly complicated, users can not use the search engine to accurately find effective information. But the emergence of the recommendation system can effectively alleviate this problem. Algorithms such as collaborative filtering have a large drop in recommended quality when user ratings are sparse. We propose a novel cross attentive collaborative filtering model. Our model can learn more about the potential relationships between users, items and ratings. To validate the validity of the algorithm model, experiments were performed on public datasets. The experimental results show that the mae and rmse index have decreased.

An empirical study of alternating least squares collaborative filtering recommendation for Movielens on Apache Hadoop and Spark

In recent years, both consumers and businesses have faced the problem of information explosion, and the recommendation system provides a possible solution. This study implements a movie recommendation system that provides recommendations to consumers in an effort to increase consumer spending while reducing the time between film selection. This study is a prototype of collaborative filtering recommendation system based on Alternating Least Squares (ALS) algorithm. The advantage of collaborative filtering is that it avoids possible violations of the Personal Data Protection Act and reduces the possibility of errors due to poor quality of personal data. Our research improves the ALS's limited scalability by using a platform that combines Spark with Hadoop Yarn and uses this combination to calculate movie recommendations and store data separately. Based on the results of this study, our proposed system architecture provides recommendations with satisfactory accuracy while maintaining acceptable computational time with limited resources.