Initial Profile Generation in Recommender Systems Using Pairwise Comparison

In recent years, multidimensional data is becoming increasingly popular in recommender systems. For example, social tagging systems encourage users to employ user-defined keywords to help manage content in a personalized way. Recommender systems built upon social tagging systems utilize social tagging data to improve recommendation accuracy. Context-aware recommender systems incorporate context information (e.g. time, location, weather, etc.) into recommendation models to make recommendations more personalized. User profiles play a crucial role in recommender systems because user profiles provide the information about users’ information needs based on which personalized recommendations can be generated. Existing user profiling techniques for collaborative filtering (CF) recommender systems based on multidimensional data mostly analyze data through splitting the multidimensional relations into lower-dimensional relations. However, this leads to the loss of multidimensionality in user-item interactions. A different type of methods for addressing multidimensional data is factorization techniques such as tensor factorization (TF), which discover holistic latent relations between all dimensions in multidimensional data. However, TF models are not constructed to analyze any particular dimension of multidimensional data. But for many application domains, certain dimensions are of particular interest such as the user dimension in recommender systems. However, user profiles are not explicitly reflected in TF models. In this paper, we propose a unified approach to profile any particular dimension in multidimensional data based on partially-decomposed Tucker models (TM). We prove that the proposed profiling approach is a compatible higher-order extension of the matrix factorization based approaches. Moreover, we prove that the proposed TM-based profiles can be updated incrementally with new data without reconstructing the entire profiles. We integrate the proposed TF-based user/item profiling approach into neighborhood-based collaborative filtering recommenders for making top-N item recommendations. Extensive experiments have been conducted using real-world social tagging datasets to demonstrate that the CF recommenders integrated with the proposed user/item profiles outperform state-of-the-art CF recommendation approaches in terms of recommendation accuracy.

Recently, online shopping has become very prominent. Furthermore, most e-commerce platforms use recommender systems to suggest products to the user for potential purchase to help enhance the user experience. Existing recommender models determine the correlation between user profiles and product information to make product recommendations. Although current recommender systems can accurately suggest products, there is a lack of transparency in terms of how the recommendations are made. As a result, users may not trust the suggestions made by what could be perceived as a black box. To address this problem, we have designed a deep recommender system that discovers user preferences and provides the reasons behind the recommendations it makes by examining clues such as user reviews and item descriptions. We adopt the attention mechanism to enhance text analysis capability, which not only enables accurate recommendations but also shows the relationship between products and user profiles. Furthermore, we conduct a case study in which we utilize feature weights to present the reasons to the user by observing attention maps. The results of our experiment show that our model can consistently provide accurate recommendations and justifications to the user. Moreover, our model can handle new items and users, that is, data that is not part of the initial training dataset. Our recommender system is a novel solution that leverages the advantages of content-based and collaborative filtering.

Recommender systems can improve the quality of life in smart cities by presenting personalized services to the community. Such systems maintain a database of user profiles for producing recommendations for a specific user. The collaborative filtering (CF) approach used in these systems has become a benchmark approach for generating recommendations for interested users because it can provide “out of the box” solutions. These CF-based approaches first construct a user–item rating matrix and then exploit similarity methods. These approaches suffer from scalability, sparsity, and cold user conditions, which consequently result in the poor recommendation accuracy of these systems. To enhance the accuracy of recommender systems, social trust can play a vital role because people tend to interact with a system or respond positively to recommendations that originate from their social trustworthy friends. The proposed unified approach of this article uses explicit trust, implicit trust, and user preference similarity to create a unified rating profile for the target user to produce more powerful and accurate recommendations. The proposed unified approach also enhances the recommendation performance of CF-based recommender systems when only a limited set of ratings is available. Experiments are performed on three publicly available datasets which are FilmTrust, CiaoDVD, and Epinions. Comparison of obtained results is made with traditional similarity measures as well as up-to-date trust-based approaches. The results show that the proposed unified approach is superior to existing approaches in terms of both predictive and classification-based accuracy measures.

Recommender system is the system which, by using automated information filtering technology, recommends products or services to the customers who are likely to be interested in. Those systems are widely used in many different Web retailers such as Amazon.com, Netfix.com, and CDNow.com. Various recommender systems have been developed. Among them, Collaborative Filtering (CF) has been known as the most successful and commonly used approach. CF identifies customers whose tastes are similar to those of a given customer, and recommends items those customers have liked in the past. Numerous CF algorithms have been developed to increase the performance of recommender systems. However, the relative performances of CF algorithms are known to be domain and data dependent. It is very time-consuming and expensive to implement and launce a CF recommender system, and also the system unsuited for the given domain provides customers with poor quality recommendations that make them easily annoyed. Therefore, predicting in advance whether the performance of CF recommender system is acceptable or not is practically important and needed. In this study, we propose a decision making guideline which helps decide whether CF is adoptable for a given application with certain transaction data characteristics. Several previous studies reported that sparsity, gray sheep, cold-start, coverage, and serendipity could affect the performance of CF, but the theoretical and empirical justification of such factors is lacking. Recently there are many studies paying attention to Social Network Analysis (SNA) as a method to analyze social relationships among people. SNA is a method to measure and visualize the linkage structure and status focusing on interaction among objects within communication group. CF analyzes the similarity among previous ratings or purchases of each customer, finds the relationships among the customers who have similarities, and then uses the relationships for recommendations. Thus CF can be modeled as a social network in which customers are nodes and purchase relationships between customers are links. Under the assumption that SNA could facilitate an exploration of the topological properties of the network structure that are implicit in transaction data for CF recommendations, we focus on density, clustering coefficient, and centralization which are ones of the most commonly used measures to capture topological properties of the social network structure. While network density, expressed as a proportion of the maximum possible number of links, captures the density of the whole network, the clustering coefficient captures the degree to which the overall network contains localized pockets of dense connectivity. Centralization reflects the extent to which connections are concentrated in a small number of nodes rather than distributed equally among all nodes. We explore how these SNA measures affect the performance of CF performance and how they interact to each other. Our experiments used sales transaction data from H department store, one of the well?known department stores in Korea. Total 396 data set were sampled to construct various types of social networks. The dependant variable measuring process consists of three steps; analysis of customer similarities, construction of a social network, and analysis of social network patterns. We used UCINET 6.0 for SNA. The experiments conducted the 3-way ANOVA which employs three SNA measures as dependant variables, and the recommendation accuracy measured by F1-measure as an independent variable. The experiments report that 1) each of three SNA measures affects the recommendation accuracy, 2) the density's effect to the performance overrides those of clustering coefficient and centralization (i.e., CF adoption is not a good decision if the density is low), and 3) however though the density is low, the performance of CF is comparatively good when the clustering coefficient is low. We expect that these experiment results help firms decide whether CF recommender system is adoptable for their business domain with certain transaction data characteristics.

Collaborative Filtering (CF), as one of the most popular approaches, is widely employed in recommender systems but suffers from the cold-start problem, where interactions are very limited for new users in the system. To deal with this issue, previous work has largely focused on utilizing various auxiliary information such as user profiles and social relationships to infer user preferences. However, the auxiliary information is not always available due to reasons such as user privacy concerns, making the CF approaches have to count on the limited interactions. Moreover, real-world situations require both accurate and quick recommendations for newly arrived users dynamically. Therefore, it is of critical importance to enable fast learning for new users during the training time of CF models. In this paper, we present a novel learning paradigm, named MetaCF, to learn an accurate CF model that makes fast adaptation on new users with limited interactions. Inspired by meta-learning, MetaCF treats the fast adaptation on a new user as a task and aims to learn a suitable model for initializing the adaption. To pursue a well-generalized model, MetaCF is equipped with a Dynamic Subgraph Sampling that accounts for the dynamic arrival of new users by dynamically generating representative adaptation tasks for existing users. Moreover, to stabilize the adaption procedure that faces the shortage of training samples, MetaCF further optimizes the learning rates for adaption in a fine-grained manner. MetaCF is applicable to any differentiable CF-based models where we demonstrate it on two representative ones, FISM [1] and NGCF [2]. Extensive experiments on three datasets validate the effectiveness of the proposed framework, which significantly outperforms state-of-the-art baselines by a large margin in the cold-start scenario where user-item interactions are limited.

Recommender Systems play a very important role in online services nowadays, with the huge repository of content being generated every day. While these systems are used on large scales, they still face a lot of problems. The item cold start is one of the most famous problems in recommender systems. While systems can develop an initial image of new user profiles through various methods, new items face a problem surfacing into user recommendations. This is because they have low to almost zero interactions, and most recommender systems use methods like collaborative filtering to share similar items to similar user profiles. This is an important problem to be dealt with. Otherwise, recommendation feeds get saturated with already popular content, which harms new creators on platforms that might be creating innovative content. We propose a hybrid recommender system: EASE'Em, that seeks to solve this problem by generating cold start items to users based on the item similarities and user profiles. Our model first uses an AutoEncoder recommendation system to fill a sparse user interaction matrix to develop a better understanding of user profiles. We then project item attributes and user profiles into a common low dimensional space and then use matrix factorization techniques to generate top-N recommendations for each user present in the set. We then compare our model generations with standard item-based recommender models on a subset of the Million Song Dataset and compare their results.

Recommender systems usually suggest items by exploiting all the previous interactions of the users with a system (e.g., in order to decide the movies to recommend to a user, all the movies she previously purchased are considered). This canonical approach sometimes could lead to wrong results due to several factors, such as a change in user preferences over time, or the use of her account by third parties. This kind of incoherence in the user profiles defines a lower bound on the error the recommender systems may achieve when they generate suggestions for a user, an aspect known in literature as magic barrier. This paper proposes a novel dynamic coherence-based approach to define the user profile used in the recommendation process. The main aim is to identify and remove, from the previously evaluated items, those not semantically adherent to the others, in order to make a user profile as close as possible to the user's real preferences, solving the aforementioned problems. Moreover, reshaping the user profile in such a way leads to great advantages in terms of computational complexity, since the number of items considered during the recommendation process is highly reduced. The performed experiments show the effectiveness of our approach to remove the incoherent items from a user profile, increasing the recommendation accuracy.

User profiling is the process of collecting information about a user in order to construct their profile. The information in a user profile may include various attributes of a user such as geographical location, academic and professional background, membership in groups, interests, preferences, opinions, etc. Big data techniques enable collecting accurate and rich information for user profiles, in particular due to their ability to process unstructured as well as structured information in high volumes from multiple sources. Accurate and rich user profiles are important for applications such as recommender systems, which try to predict elements that a user has not yet considered but may find useful. The information contained in user profiles is personal and thus there are privacy issues related to user profiling. In this position paper, we discuss user profiling with big data techniques and the associated privacy challenges. We also discuss the ongoing EU-funded EEXCESS project as a concrete example of constructing user profiles with big data techniques and the approaches being considered for preserving user privacy.

The most popular Recommender systems (RSs) employ Collaborative Filtering (CF) algorithms where users explicitly rate items. Based on these ratings, a user-item rating matrix is generated and used to select the items to be recommended for a target user. An important step in this process is to determine the neighborhood of a target user, i.e. a set of users who rate items similarly to this user. One of the limitations of CF is precisely the need of rating data provided voluntarily by users. The lack of interest of users to provide this kind of information increases the sparsity problem of the ratings matrix. In this paper, we propose the use of implicit feedback for neighbors selection to alleviate the sparsity problem in CF-based RSs. In this proposal, user profiles are built based on the characteristics of items that have been accessed or purchased, and not necessarily rated by the users. This user profile is used exclusively to the neighborhoods formation, which considers not how they have rated items, but by the characteristics of the items that they have accessed or purchased. Our technique was implemented with Apache Mahout Framework and evaluated across experiments in the domain of movies by using a dataset from Movielens project. The results demonstrated that our technique produces better quality recommendations when compared to the classic CF mainly in presence of sparsity of rating data.

User profile release in recommendation systems can apply the user profile perturbation technique to protect user privacy, in which each user sends a perturbed user profile such as the a list of clicked items to receive a recommendation service from a server. The perturbation policy such as the privacy budget determines the recommendation quality and the privacy level, while its optimization usually depends on the known attack model, which is rarely known by the users. In this paper, we propose a reinforcement learning based user profile perturbation scheme that applies differential privacy to protect user privacy for recommendation systems. According to reinforcement learning, the privacy budget to perturb the released user profile depends on the features of the actual user profiles and the released user profiles, and the estimated user privacy level. This scheme enables a user to optimize his or her perturbation policy in terms of both the user privacy level and the received recommendation quality without being aware of the attack model. We evaluate the computational complexity of this scheme and analyze a case study, a privacy aware movie recommendation system. Simulation results show that this scheme improves user privacy protection for a given level of recommendation quality compared with a benchmark profile perturbation scheme.

Web based commercial recommender systems (RS) can help users to make decisions about which product to purchase from the vast amount of products available on the Internet. Currently, many commercial recommender systems are developed for recommending frequently purchased products where a large amount of explicit ratings or purchase history data is available to predict user preferences. However, for products that are infrequently purchased by users, it is difficult to collect such data and, thus, user profiling becomes a major challenge for recommending these kinds of products. This paper proposes a recommendation approach for infrequently purchased products based on user opinions and navigation data. User opinion data, which is collected from product review data, is used to generate product profiles and user navigation data is used to generate user profiles, both of which are used for recommending products that best satisfy the users' needs. Experiments conducted on real e-commerce data show that the proposed approach, named, Adaptive Collaborative Filtering (ACF), which utilizes user and product profiles, outperforms the Query Expansion (QE) approach that only utilizes product profiles to recommend products. The ACF also performs better than the Basic Search (BS) approach, which is widely applied by the current e-commerce applications.

Collaborative tags labeled by users in social tagging systems contain rich information about individual preference and resource content, which can further improve the performance of personalized recommender system. In this paper, a hybrid method that combines the collaborative filtering with graph-based interest propagation is proposed for movie recommendation. In the proposed method, both user and movie profiles are constructed by tags. The top-k similar users are collected by user-user interest which is calculated by the user profiles, and then a user-movie bipartite graph is constructed according to the top-k users and candidate movies. We utilize the user and movie profiles to calculate user-movie interest, and the interest is propagated in the graph until converged or the max iterations are reached. Lastly, the top-n movies in the graph are recommended to the initial user. Experimental results on MovieLens dataset demonstrate that our proposed method can achieve better performance than several baselines for the movie recommendation problem.

In the open social networks, the analysis of user data after the injection attack has a great impact on the recommendation system. K-Nearest Neighbor-based collaborative filtering algorithms are very vulnerable to this attack. Another recommendation algorithm based on probabilistic latent semantic analysis has relatively accurate recommendation, but it is not very stable and robust against attacks on the overall user data of the recommendation system. Here is used to DeepWalk the user network processing, while taking advantage of the user profile feature time series to consider the user’s behavior over time, the algorithm also analyzes the stability and robustness of DeepWalk and user profile. The results show that especially the DeepWalk-based approach can achieve comparable recommendation accuracy.

Recommender Systems is a special type of information filtering system which has become important in the information overloaded and strategic decision making environment. Recommender System is used to produce meaningful suggestions about new items for particular consumers. These recommendations may be based on the user profile or item ratings, facilitate the users to make decisions in multiple contexts, such as what items to buy, what online news to read or what music to listen. Recommender Systems helps their founders to increase profits by recommending items and attracting new consumers. Collaborative filtering technique recommends items basis of conclusion of opinions about various products by users of similar profile to the active user. This technique requires user-items-ratings matrix. Although this is the most mature and commonly implemented technique, it faces major problem of Data Sparsity problem. Sparsity Problem occurs as a result of lack of enough information when only a few of the total number of items are rated by the users. This produces a sparse user item matrix leads to weak recommendations. This paper presents a recommender system using collaborative filtering implemented with RapidMiner tool. The proposed recommendation system is designed with users’ similarity calculated by Sequence and Set Similarity Measure (S3M) with utilizing similarity upper approximation and a Singular Value Decomposition (SVD) model based technique used for recommending ratings for removing sparsity.

One of the prevalent research challenges in the field of recommender system is to do better user profiling. There are some advanced user profiling techniques found in the literature to achieve the same. User profiling aims to understand the user well and as a result recommending the most relevant items to the user, where relevant means items returned as a result of intelligent techniques from various fields, mainly from data mining. This work is an attempt to answer the question “who understands a user the most?” The three obvious answers are Recommender System’s high end approaches (e.g. data mining and statistical approaches), neighbors of the user or the user herself. The correct answer would be the last one, which is a user knows herself the best. In this direction, we propose to make users empowered and responsible for registering their preferences and sharing the same at their discretion. More personalized solutions can be offered when a user tells what she prefers and can contribute explicitly to the recommendation system’s results generation. When a user is given the handle to communicate her preferences to the recommender system, more personalized recommendations can be given which are not only relevant (as tested by sophisticated evaluation matrices for recommender systems) but also plays wonder to users’ satisfaction.