Biblio

A recommender system aims to suggest the most relevant items to users based on their personal data. However, data privacy is a growing concern for anyone. Secure recommender system is a research direction to preserve user privacy while maintaining as high performance as possible. The most recent strategy is to use Federated Learning, a machine learning technique for privacy-preserving distributed training. In Federated Learning, a subset of users will be selected for training model using data at local systems, the server will securely aggregate the computing result from local models to generate a global model, finally that model will give recommendations to users. In this paper, we present a novel algorithm to train Collaborative Filtering recommender system specialized for the ranking task in Federated Learning setting, where the goal is to protect user interaction information (i.e., implicit feedback). Specifically, with the help of the algorithm, the recommender system will be trained by Neural Collaborative Filtering, one of the state-of-the-art matrix factorization methods and Bayesian Personalized Ranking, the most common pairwise approach. In contrast to existing approaches which protect user privacy by requiring users to download/upload the information associated with all interactions that they can possibly interact with in order to perform training, the algorithm can protect user privacy at low communication cost, where users only need to obtain/transfer the information related to a small number of interactions per training iteration. Above all, through extensive experiments, the algorithm has demonstrated to utilize user data more efficient than the most recent research called FedeRank, while ensuring that user privacy is still preserved.

In order to solve the problem that collaborative filtering recommendation algorithm completely depends on the interactive behavior information of users while ignoring the correlation information between items, this paper introduces a link prediction algorithm based on knowledge graph to integrate ItemCF algorithm. Through the linear weighted fusion of the item similarity matrix obtained by the ItemCF algorithm and the item similarity matrix obtained by the link prediction algorithm, the new fusion matrix is then introduced into ItemCF algorithm. The MovieLens-1M data set is used to verify the KGLP-ItemCF model proposed in this paper, and the experimental results show that the KGLP-ItemCF model effectively improves the precision, recall rate and F1 value. KGLP-ItemCF model effectively solves the problems of sparse data and over-reliance on user interaction information by introducing knowledge graph into ItemCF algorithm.

With the proliferation of false redundant information on various e-commerce platforms, ineffective recommendations and other untrustworthy behaviors have seriously hindered the healthy development of e-commerce platforms. Modern recommendation systems often use side information to alleviate these problems and also increase prediction accuracy. One such piece of side information, which has been widely investigated, is trust. However, it is difficult to obtain explicit trust relationship data, so researchers infer trust values from other methods, such as the user-to-item relationship. In this paper, addressing the problems, we proposed a novel trust-based recommender model called UITrust, which uses user-item relationship value to improve prediction accuracy. With the improvement the traditional similarity measures by employing the entropies of user and item history ratings to reflect the global rating behavior on both. We evaluate the proposed model using two real-world datasets. The proposed model performs significantly better than the baseline methods. Also, we can use the UITrust to alleviate the sparsity problem associated with correlation-based similarity. In addition to that, the proposed model has a better computational complexity for making predictions than the k-nearest neighbor (kNN) method.

In Social Aware Network (SAN) model, the elementary actions focus on investigating the attributes and behaviors of the customer. This analysis of customer attributes facilitate in the design of highly active and improved protocols. In specific, the recommender systems are highly vulnerable to the shilling attack. The recommender system provides the solution to solve the issues like information overload. Collaborative filtering based recommender systems are susceptible to shilling attack known as profile injection attacks. In the shilling attack, the malicious users bias the output of the system's recommendations by adding the fake profiles. The attacker exploits the customer reviews, customer ratings and fake data for the processing of recommendation level. It is essential to detect the shilling attack in the network for sustaining the reliability and fairness of the recommender systems. This article reviews the most prominent issues and challenges of shilling attack. This paper presents the literature survey which is contributed in focusing of shilling attack and also describes the merits and demerits with its evaluation metrics like attack detection accuracy, precision and recall along with different datasets used for identifying the shilling attack in SAN network.

In the last few years, cluster ensembles have emerged as powerful techniques that integrate multiple clustering methods into recommender systems. Such integration leads to improving the performance, quality and the accuracy of the generated recommendations. This paper proposes a novel recommender system based on a cluster ensemble technique for big data. The proposed system incorporates the collaborative filtering recommendation technique and the cluster ensemble to improve the system performance. Besides, it integrates the Expectation-Maximization method and the HyperGraph Partitioning Algorithm to generate new recommendations and enhance the overall accuracy. We use two real-world datasets to evaluate our system: TED Talks and MovieLens. The experimental results show that the proposed system outperforms the traditional methods that utilize single clustering techniques in terms of recommendation quality and predictive accuracy. Most importantly, the results indicate that the proposed system provides the highest precision, recall, accuracy, F1, and the lowest Root Mean Square Error regardless of the used similarity strategy.

Recommender systems help people in finding a particular item based on their preference from a wide range of products in online shopping rapidly. One of the most popular models of recommendation systems is the Collaborative Filtering Recommendation System (CFRS) that recommend the top-K items to active user based on peer grouping user ratings. The implementation of CFRS is easy and it can easily be attacked by fake users and affect the recommendation. Fake users create a fake profile to attack the RS and change the output of it. Different attack types with different features and attacking methods exist in which decrease the accuracy. It is important to detect fake users, remove their rating from rating matrix and recognize the items has been attacked. In the recent years, many algorithms have been proposed to detect the attackers but first, researchers have to inject the attack type into their dataset and then evaluate their proposed approach. The purpose of this article is to develop a tool to inject the different attack types to datasets. Proposed tool constructs a new dataset containing the fake users therefore researchers can use it for evaluating their proposed attack detection methods. Researchers could choose the attack type and the size of attack with a user interface of our proposed tool easily.

Recommender systems (RSs) are intelligent systems to help e-commerce users to find their preferred items among millions of available items by considering the profiles of both users and items. These systems need to predict the unknown ratings and then recommend a set of high rated items. Among the others, Collaborative Filtering (CF) is a successful recommendation approach and has been utilized in many real-world systems. CF methods seek to predict missing ratings by considering the preferences of those users who are similar to the target user. A major task in Collaborative Filtering is to identify an accurate set of users and employing them in the rating prediction process. Most of the CF-based methods suffer from the cold-start issue which arising from an insufficient number of ratings in the prediction process. This is due to the fact that users only comment on a few items and thus CF methods faced with a sparse user-item matrix. To tackle this issue, a new collaborative filtering method is proposed that has a trust-aware strategy. The proposed method employs the trust relationships of users as additional information to help the CF tackle the cold-start issue. To this end, the proposed integrated trust relationships in the prediction process by using the Ant Colony Optimization (ACO). The proposed method has four main steps. The aim of the first step is ranking users based on their similarities to the target user. This step uses trust relationships and the available rating values in its process. Then in the second step, graph clustering methods are used to cluster the trust graph to group similar users. In the third step, the users are weighted based on their similarities to the target users. To this end, an ACO process is employed on the users' graph. Finally, those of top users with high similarity to the target user are used in the rating prediction process. The superiority of our method has been shown in the experimental results in comparison with well-known and state-of-the-art methods.

Collaborative Filtering (CF) is a popular recommendation system that makes recommendations based on similar users' preferences. Though it is widely used, CF is prone to Shilling/Profile Injection attacks, where fake profiles are injected into the CF system to alter its outcome. Most of the existing shilling attacks do not work on online systems and cannot be efficiently implemented in real-world applications. In this paper, we introduce an efficient Multi-Armed-Bandit-based reinforcement learning method to practically execute online shilling attacks. Our method works by reducing the uncertainty associated with the item selection process and finds the most optimal items to enhance attack reach. Such practical online attacks open new avenues for research in building more robust recommender systems. We treat the recommender system as a black box, making our method effective irrespective of the type of CF used. Finally, we also experimentally test our approach against popular state-of-the-art shilling attacks.

In this paper, differential privacy protection method is applied to matrix factorization method that used to solve the recommendation problem. For centralized recommendation scenarios, a collaborative filtering recommendation model based on matrix factorization is established, and a matrix factorization mechanism satisfying ε-differential privacy is proposed. Firstly, the potential characteristic matrix of users and projects is constructed. Secondly, noise is added to the matrix by the method of target disturbance, which satisfies the differential privacy constraint, then the noise matrix factorization model is obtained. The parameters of the model are obtained by the stochastic gradient descent algorithm. Finally, the differential privacy matrix factorization model is used for score prediction. The effectiveness of the algorithm is evaluated on the public datasets including Movielens and Netflix. The experimental results show that compared with the existing typical recommendation methods, the new matrix factorization method with privacy protection can recommend within a certain range of recommendation accuracy loss while protecting the users' privacy information.

The rapid development of the Internet and ecommerce has brought a lot of convenience to people's life. Personalized recommendation technology provides users with services that they may be interested according to users' information such as personal characteristics and historical behaviors. The research of personalized recommendation has been a hot point of data mining and social networks. In this paper, we focus on resolving the problem of data sparsity based on users' rating data and social network information, introduce a set of new measures for social trust and propose a novel personalized recommendation algorithm based on matrix factorization combining trust relevancy. Our experiments were performed on the Dianping datasets. The results show that our algorithm outperforms traditional approaches in terms of accuracy and stability.

Matrix factorization (MF) has been proved to be an effective approach to build a successful recommender system. However, most current MF-based recommenders cannot obtain high prediction accuracy due to the sparseness of user-item matrix. Moreover, these methods suffer from the scalability issues when applying on large-scale real-world tasks. To tackle these issues, in this paper a social regularization method called TrustRSNMF is proposed that incorporates the social trust information of users in nonnegative matrix factorization framework. The proposed method integrates trust statements along with user-item ratings as an additional information source into the recommendation model to deal with the data sparsity and cold-start issues. In order to evaluate the effectiveness of the proposed method, a number of experiments are performed on two real-world datasets. The obtained results demonstrate significant improvements of the proposed method compared to state-of-the-art recommendation methods.

In order to improve the accuracy of similarity, an improved collaborative filtering algorithm based on trust and information entropy is proposed in this paper. Firstly, the direct trust between the users is determined by the user's rating to explore the potential trust relationship of the users. The time decay function is introduced to realize the dynamic portrayal of the user's interest decays over time. Secondly, the direct trust and the indirect trust are combined to obtain the overall trust which is weighted with the Pearson similarity to obtain the trust similarity. Then, the information entropy theory is introduced to calculate the similarity based on weighted information entropy. At last, the trust similarity and the similarity based on weighted information entropy are weighted to obtain the similarity combing trust and information entropy which is used to predicted the rating of the target user and create the recommendation. The simulation shows that the improved algorithm has a higher accuracy of recommendation and can provide more accurate and reliable recommendation service.

The security of user personal information on cloud computing is an important issue for the recommendation system. In order to provide high quality recommendation services, privacy of user is often obtained by untrusted recommendation systems. At the same time, malicious attacks often use the recommendation results to try to guess the private data of user. This paper proposes a hybrid algorithm based on NMF and random perturbation technology, which implements the recommendation system and solves the protection problem of user privacy data in the recommendation process on cloud computing. Compared with the privacy protection algorithm of SVD, the elements of the matrix after the decomposition of the new algorithm are non-negative elements, avoiding the meaninglessness of negative numbers in the matrix formed by texts, images, etc., and it has a good explanation for the local characteristics of things. Experiments show that the new algorithm can produce recommendation results with certain accuracy under the premise of protecting users' personal privacy on cloud computing.

The condition of road surface has a significant role in land transportation. Due to poor road conditions, the logistics and supply chain industry face a drastic loss in their business. Unmaintained roads can cause damage to goods and accidents. The existing routing techniques do not consider factors like shock, temperature and tilt of goods etc. but these factors have to be considered for the logistics and supply chain industry. This paper proposes a recommender system which target management of goods in logistics. A 3 axis accelerometer is used to measure the road surface conditions. The pothole location is obtained using Global Positioning System (GPS). Using these details a hybrid recommender system is built. Hybrid recommender system combines multiple recommendation techniques to develop an effective recommender system. Here content-based and collaborative-based techniques is combined to build a hybrid recommender system. One of the popular Multiple Criteria Decision Making (MCDM) method, The Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) is used for content based filtering and normalised Euclidean distance and KNN algorithm is used for collaborative filtering. The best route recommended by the system will be displayed to the user using a map application.

Collaborative filtering (CF) is a successful and hence most widely used technique for recommender systems. However, it is vulnerable to shilling attack due to its open nature, which results in generating biased or false recommendations for users. In literature, segment attack (push attack) has been widely studied and investigated while rare studies have been performed on nuke attack, to the best of our knowledge. Further, the robustness of binary collaborative filtering and hybrid approach has not been investigated against segment-focused attack. In this paper, from the perspective of robustness, binary collaborative filtering, hybrid approach, stand-alone rating user-based, and stand-alone rating item- based recommendation have been evaluated against segment attack on a large dataset (100K ratings) which is found to be more successful as it attacks target set of items. With an aim to find an approach which reflects a higher accuracy in recommending items and is less vulnerable to segment-based attack, the possibility of any relationship between accuracy and vulnerability of six CF approaches were studied. Such an approach needs to be re-examined by the researchers marking the future of recommender system (RS). Experimental results show negligible positive correlation between accuracy and vulnerability of techniques. Publicly available dataset namely MovieLens was used for conducting experiments. Robustness and accuracy of CF techniques were calculated using prediction shift and F-measure, respectively.

Spam is a genuine and irritating issue for quite a longtime. Despite the fact that a lot of arrangements have been advanced, there still remains a considerable measure to be advanced in separating spam messages all the more proficiently. These days a noteworthy issue in spam separating also as content characterization in common dialect handling is the colossal size of vector space because of the various element terms, which is normally the reason for broad figuring and moderate order. Extracting semantic implications from the substance of writings and utilizing these as highlight terms to develop the vector space, rather than utilizing words as highlight terms in convention ways, could decrease the component of vectors viably and advance the characterization in the meantime. In spite of the fact that there are a wide range of techniques to square spam messages, a large portion of program designers just mean to square spam messages from being conveyed to their customers. In this paper, we present an effective way to deal with keep spam messages from being exchanged.In this work, a Collaborative filtering approach with semantics-based text classification technology was proposed and the related feature terms were selected from the semantic meanings of the text content.

In this paper, we propose a hybrid collaborative filtering recommendation algorithm based on heuristic similarity and trust measure, in order to alleviate the problem of data sparsity, cold start and trust measure. Firstly, a new similarity measure is implemented by weighted fusion of multiple similarity influence factors obtained from the rating matrix, so that the similarity measure becomes more accurate. Then, a user trust relationship computing model is implemented by constructing the user's trust network based on the trust propagation theory. On this basis, a SIMT collaborative filtering algorithm is designed which integrates trust and similarity instead of the similarity in traditional collaborative filtering algorithm. Further, an improved K nearest neighbor recommendation based on clustering algorithm is implemented for generation of a better recommendation list. Finally, a comparative experiment on FilmTrust dataset shows that the proposed algorithm has improved the quality and accuracy of recommendation, thus overcome the problem of data sparsity, cold start and trust measure to a certain extent.

The risk of cyber-attacks exploiting vulnerable organisations has increased significantly over the past several years. These attacks may combine to exploit a vulnerability breach within a system's protection strategy, which has the potential for loss, damage or destruction of assets. Consequently, every vulnerability has an accompanying risk, which is defined as the "intersection of assets, threats, and vulnerabilities" [1]. This research project aims to experimentally compare the similarity-based ranking of cyber security information utilising a recommendation environment. The Memory-Based Collaborative Filtering technique was employed, specifically the User-Based and Item-Based approaches. These systems utilised information from the National Vulnerability Database, specifically for the identification and similarity-based ranking of cyber-security vulnerability information, relating to hardware and software applications. Experiments were performed using the Item-Based technique, to identify the optimum system parameters, evaluated through the AUC evaluation metric. Once identified, the Item-Based technique was compared with the User-Based technique which utilised the parameters identified from the previous experiments. During these experiments, the Pearson's Correlation Coefficient and the Cosine similarity measure was used. From these experiments, it was identified that utilised the Item-Based technique which employed the Cosine similarity measure, an AUC evaluation metric of 0.80225 was achieved.

CFRS (Collaborative Filtering Recommendation System) is one of the most widely used individualized recommendation systems. However, CFRS is susceptible to shilling attacks based on profile injection. The current research on shilling attack mainly focuses on the recognition of false user profiles, but these methods depend on the specific attack models and the computational cost is huge. From the view of item, some abnormal item detection methods are proposed which are independent of attack models and overcome the defects of user profiles model, but its detection rate, false alarm rate and time overhead need to be further improved. In order to solve these problems, it proposes an abnormal item detection method based on time window merging. This method first uses the small window to partition rating time series, and determine whether the window is suspicious in terms of the number of abnormal ratings within it. Then, the suspicious small windows are merged to form suspicious intervals. We use the rating distribution characteristics RAR (Ratio of Abnormal Rating), ATIAR (Average Time Interval of Abnormal Rating), DAR(Deviation of Abnormal Rating) and DTIAR (Deviation of Time Interval of Abnormal Rating) in the suspicious intervals to determine whether the item is subject to attacks. Experiment results on the MovieLens 100K data set show that the method has a high detection rate and a low false alarm rate.

Collaborative filtering (CF) recommender system has been widely used for its well performing in personalized recommendation, but CF recommender system is vulnerable to shilling attacks in which shilling attack profiles are injected into the system by attackers to affect recommendations. Design robust recommender system and propose attack detection methods are the main research direction to handle shilling attacks, among which unsupervised PCA is particularly effective in experiment, but if we have no information about the number of shilling attack profiles, the unsupervised PCA will be suffered. In this paper, a new unsupervised detection method which combine PCA and data complexity has been proposed to detect shilling attacks. In the proposed method, PCA is used to select suspected attack profiles, and data complexity is used to pick out the authentic profiles from suspected attack profiles. Compared with the traditional PCA, the proposed method could perform well and there is no need to determine the number of shilling attack profiles in advance.

This paper evaluates two algorithms, BLIP and JLT, for creating differentially private data sketches of user profiles, in terms of their ability to protect a kNN collaborative filtering algorithm from an inference attack by third-parties. The transformed user profiles are employed in a user-based top-N collaborative filtering system. For the first time, a theoretical analysis of the BLIP is carried out, to derive expressions that relate its parameters to its performance. This allows the two techniques to be fairly compared. The impact of deploying these approaches on the utility of the system—its ability to make good recommendations, and on its privacy level—the ability of third-parties to make inferences about the underlying user preferences, is examined. An active inference attack is evaluated, that consists of the injection of a number of tailored sybil profiles into the system database. User profile data of targeted users is then inferred from the recommendations made to the sybils. Although the differentially private sketches are designed to allow the transformed user profiles to be published without compromising privacy, the attack we examine does not use such information and depends only on some pre-existing knowledge of some user preferences as well as the neighbourhood size of the kNN algorithm. Our analysis therefore assesses in practical terms a relatively weak privacy attack, which is extremely simple to apply in systems that allow low-cost generation of sybils. We find that, for a given differential privacy level, the BLIP injects less noise into the system, but for a given level of noise, the JLT offers a more compact representation.

Collaborative filtering (CF) has made it possible to build personalized recommendation models leveraging the collective data of large user groups, albeit with prescribed models that cannot easily leverage the existence of known behavioral models in particular settings. In this paper, we facilitate the combination of CF with existing behavioral models by introducing Bayesian Behavioral Collaborative Filtering (BBCF). BBCF works by embedding arbitrary (black-box) probabilistic models of human behavior in a latent variable Bayesian framework capable of collectively leveraging behavioral models trained on all users for personalized recommendation. There are three key advantages of BBCF compared to traditional CF and non-CF methods: (1) BBCF can leverage highly specialized behavioral models for specific CF use cases that may outperform existing generic models used in standard CF, (2) the behavioral models used in BBCF may offer enhanced intepretability and explainability compared to generic CF methods, and (3) compared to non-CF methods that would train a behavioral model per specific user and thus may suffer when individual user data is limited, BBCF leverages the data of all users thus enabling strong performance across the data availability spectrum including the near cold-start case. Experimentally, we compare BBCF to individual and global behavioral models as well as CF techniques; our evaluation domains span sequential and non-sequential tasks with a range of behavioral models for individual users, tasks, or goal-oriented behavior. Our results demonstrate that BBCF is competitive if not better than existing methods while still offering the interpretability and explainability benefits intrinsic to many behavioral models.

Many recent advances of network caching focus on i) more effectively modeling the preferences of a regional user group to different web contents, and ii) reducing the cost of content delivery by storing the most popular contents in regional caches. However, the context under which the users interact with the network system usually causes tremendous variations in a user group's preferences on the contents. To effectively leverage such contextual information for more efficient network caching, we propose a novel mechanism to incorporate context-aware collaborative filtering into demand-driven caching. By differentiating the characterization of user interests based on a priori contexts, our approach seeks to enhance the cache performance with a more dynamic and fine-grained cache allocation process. In particular, our approach is general and adapts to various types of context information. Our evaluation shows that this new approach significantly outperforms previous non-demand-driven caching strategies by offering much higher cached content rate, especially when utilizing the contextual information.

In an Internet of Things (IOT) network, each node (device) provides and requires services and with the growth in IOT, the number of nodes providing the same service have also increased, thus creating a problem of selecting one reliable service from among many providers. In this paper, we propose a scalable graph-based collaborative filtering recommendation algorithm, improved using trust to solve service selection problem, which can scale to match the growth in IOT unlike a central recommender which fails. Using this recommender, a node can predict its ratings for the nodes that are providing the required service and then select the best rated service provider.

Social networking sites such as Flickr, YouTube, Facebook, etc. contain huge amount of user contributed data for a variety of real-world events. We describe an unsupervised approach to the problem of automatically detecting subgroups of people holding similar tastes or either taste. Item or taste tags play an important role in detecting group or subgroup, if two or more persons share the same opinion on the item or taste, they tend to use similar content. We consider the latter to be an implicit attitude. In this paper, we have investigated the impact of implicit and explicit attitude in two genres of social media discussion data, more formal wikipedia discussions and a debate discussion forum that is much more informal. Experimental results strongly suggest that implicit attitude is an important complement for explicit attitudes (expressed via sentiment) and it can improve the sub-group detection performance independent of genre. Here, we have proposed taste-based group, which can enhance the quality of service.