Traditional Recommender Systems Research Articles

Cross-platform dynamic goods recommendation system based on reinforcement learning and social networks

Aiming at the problems of cold start, gray sheep and sparsity of the traditional collaborative filtering recommendation system, this paper proposes a cross-platform dynamic goods recommendation system based on reinforcement learning and edge computing. First of all, this system models the current friendship relationship networks and potential friendship relationship networks, it also constructs two layers preference prediction models. Then, we consider the frequent change characteristic of social networks and shopping platforms, we design a dynamic reinforcement learning method and edge computing to learn the minimized entropy loss error. Finally, we finish the validation experiments based on the real datasets, the results show the proposed system realizes better link prediction accuracy, and using our proposed system can obtain an obvious increase in the accuracy compared to the existing of collaborative filtering recommendation systems.

Extending collaborative filtering recommendation using word embedding: A hybrid approach

SummaryCollaborative filtering recommendation systems, which analyze sets of user ratings, have been applied to various domains and have resulted in considerable improvements in the traditional recommendation system. However, they still have problems with data sparsity and cold‐start of the user ratings. To solve these problems, we present a hybrid recommendation approach by combining collaborative filtering methods and word embedding‐based content analysis. This study focuses on the movie domain, and therefore, the contents of the items are represented as a set of features such as titles, genres, directors, actors, and plots. The main aim of this paper is to understand the content of the movie plot using a word embedding to improve the measurement of similarity of each plot content to other plot content (called plot embedding). To enhance the accuracy in measuring the similarity between movies, we also consider other features such as titles, genres, directors, and actors extracted from movies. In the experiments, the movie dataset was collected by our crowdsourcing platform, which is the OMS platform. The experimental findings indicate that the proposed approach can enhance the efficiency of applied collaborative filtering recommendation systems.

A sequence-based and context modelling framework for recommendation

Since the last decade, data collection is becoming more pervasive, passive and easier to perform. This is resulting in the rise of data wherein a user performs some activities in a sequence, such as locations visited, physical activities performed, and modes of transport taken. In such cases, activities are often performed in a particular order, and each activity in turn may influence the subsequent activities to be performed. Moreover, such activities may be associated with multiple features or contexts, such as location, time, weather, etc. The order encoded in such data, along with the context, capture important information when it comes to modelling the preferences and personal habits of users. Traditional recommender systems, however, typically do not consider the order in which users perform activities and there is little work which considers both sequence and context simultaneously. In this work, a generic recommendation framework is proposed which leverages both sequences and context in user activity data for activity recommendation. To model user activities, a semantic view of the user’s past activities as a timeline of activity objects is presented. An essential step in the recommendation process is finding patterns in past activities performed which are closely aligned to the recent activities undertaken by the user. To calculate the distance between timelines, a novel two-level distance metric is presented which calculates distance with respect to the order of the activities as well as the context features associated with each activity occurrence. The efficacy of the proposed activity recommendation framework in various recommendation scenarios, is demonstrated using real-world datasets from multiple domains.

A survey of recommendation systems based on deep learning

Faced with massive amounts of data, people may not be able to choose the items they like. The recommendation system came into being, and it has achieved a breakthrough for a long time. Using deep learning can mine the hidden attributes of users' items and integrate them well, bringing new changes to the recommendation system. This article describes the deep learning-based recommendation system and the traditional recommendation system, and analyzes their advantages and disadvantages.

A face attribute based recommendation system via integrating denoising autoencoder and hash coding

Nowadays, with the rapid development of commerce, how to effectively improve the performance of an recommendation system has aroused great concern. However, traditional recommendation system requires users to log in their accounts, which brings poor user experience. This paper presents a novel recommendation system by using face recognition technologies to extract face attribute information as the input automatically. The system first obtains the user information of identity, gender, age, and then gets feedback by expression analysis. Based on the acquired face attributes, we propose to extract compact binary user features by integrating denoising autoencoder and hash coding, which can effectively improve the computing efficiency.The hash features from DAE-H-Face and DAE are further combined to enhance the representation ability. Finally, Hamming similarity-based collaborative filtering is used for recommendation. Experimental results on the MovieLens database show that the proposed recommendation method has better effectiveness and robustness. Moreover, the results also demonstrate its advantages to the cold start problem.

WITHDRAWN: Algorithm for movie recommendation system using collaborative filtering

Recommender systems are information filtering system that predicts the rating for users and items, basically from big data to recommend their likes. Movie recommendation systems provide a mechanism to assist users in classifying users with similar interests. Most traditional recommender systems lack accuracy in the case where data used in the recommendation process is sparse. This study addresses the sparsity problem and aims to eliminate it through a singular value decomposition collaborative filtering approach applied to a web-based movie recommendation system. This paper proposed a movie recommendation system whose primary objective is to suggest a recommender list through singular value decomposition collaborative filtering and cosine similarity. We enhance the model with a factorization form, which greatly reduces the model's number of parameters with a controlled complexity. This paper proposed a movie recommendation system whose primary objective is to suggest a recommender list through singular value decomposition collaborative filtering and cosine similarity. The present work improves these approaches by taking the movies' content information into account during the item similarity calculations. The proposed approach recommends the top n recommendation list of movies to users on user's interest preferences that were not already rated. Graphically shows the percentage of already viewed movies by user and movies recommended to User.

Attention neural collaboration filtering based on GRU for recommender systems

The collaborative filtering method is widely used in the traditional recommendation system. The collaborative filtering method based on matrix factorization treats the user’s preference for the item as a linear combination of the user and the item latent vectors, and cannot learn a deeper feature representation. In addition, the cold start and data sparsity remain major problems for collaborative filtering. To tackle these problems, some scholars have proposed to use deep neural network to extract text information, but did not consider the impact of long-distance dependent information and key information on their models. In this paper, we propose a neural collaborative filtering recommender method that integrates user and item auxiliary information. This method fully integrates user-item rating information, user assistance information and item text assistance information for feature extraction. First, Stacked Denoising Auto Encoder is used to extract user features, and Gated Recurrent Unit with auxiliary information is used to extract items’ latent vectors, respectively. The attention mechanism is used to learn key information when extracting text features. Second, the latent vectors learned by deep learning techniques are used in multi-layer nonlinear networks to learn more abstract and deeper feature representations to predict user preferences. According to the verification results on the MovieLens data set, the proposed model outperforms other traditional approaches and deep learning models making it state of the art.

Context-aware recommender system using trust network

Context-Aware Recommender Systems (CARS) improve traditional Recommender Systems (RS) in a wide array of domains and applications. However, CARS suffer from several inherent issues such as data sparsity and cold start. Incorporating trust into recommender systems can handle these issues. Trust-aware recommender systems use information from social networks such as trust statements, which prove another valuable information source. This paper exploits the advantages of these two systems by incorporating both trust and context information. We propose a hybrid approach: Trust based Context aware Post Filtering Approach that uses trust statements as a rich information with context compensation method of contextual post-filtering approach. Our approach utilizes the relative average difference among the context on output of trust aware collaborative filtering by incorporating explicit and implicit trust information. We also use a confidence concept to remove non-confident users from the trust network before generating prediction. The performed experiments show that the proposed approach improves upon the standard RS and outperforms recommendation approaches on real world dataset.

A two-stage personalized recommendation based on multi-objective teaching–learning-based optimization with decomposition

Due to its successful application in information filtering and knowledge retrieval systems in the era of big data, personalized recommender system plays a significant role in meeting personalized demands of people from big data and has become a hot research hotspot. Traditional recommender systems only guarantee the recommendation accuracy and the recommendation diversity will be lost. In this paper, we propose a two-stage personalized recommendation (TSPR) algorithm based on an improved collaborative filtering (ICF) and multi-objective teaching–learning-based optimization (MOTLBO/D) with decomposition. Firstly, ICF made use of each user’ preference and social neighborhood information to obtain a candidate recommendation list for each target user. In the proposed ICF method, the jump relationship between users and their unrated items was mined to describe the characteristic behavior information of users more fully. Moreover, MOTLBO/D was adopted to simultaneously optimize accuracy and diversity of recommendation list of items for each target user. In the proposed MOTLBO/D algorithm, the learner representation strategy is designed based on attributes of the recommendation problem, and the learners are updated by an improved TLBO procedure. Finally, the simulation results on two Movielens datasets show that TSPR is effective and efficient in personalized recommender systems.

Customer segmentation based on activity monitoring applications for the recommendation system

Nowadays, wearable sports performance tracking devices are gaining tremendous popularity due to intense advertising and innovative health benefits they are supposed to provide. Fashion and the opportunity to show off your results on social media are also important. On the other hand, the data recorded and shared by these devices may serve as an exciting source of customer insight for sports apparel and equipment marketers. In this study, the author proposes incorporating the data on customers’ physical activity into traditional recommendation systems. The focus was on recommendations for people who actively jog or run. The data for the recommendation system in the discussed case should characterize the specificity of the activity and be semantically related to products recommended to a specific group of people. Therefore, the analysis relies on the user activity profile and parameters of the products. The study presented in this paper introduces an analysis of an experimental dataset acquired from 210 active joggers and runners who use different sport-tracking applications. The study participants are potential customers for sports apparel and specialized types of running shoes. The groups of users were identified according to the most meaningful parameters of their workouts. Clustering and different classification models were tested with Orange software, and basic rules of division were defined. The study shows that quite an accurate user segmentation is possible and indicates its further improvement. Combining different knowledge models describing customer behavior and characteristics from various perspectives may lead to acquiring valuable and unique marketing knowledge.

An interpretable mechanism for personalized recommendation based on cross feature

The existing recommender system provides personalized recommendation service for users in online shopping, entertainment, and other activities. In order to improve the probability of users accepting the system’s recommendation service, compared with the traditional recommender system, the interpretable recommender system will give the recommendation reasons and results at the same time. In this paper, an interpretable recommendation model based on XGBoost tree is proposed to obtain comprehensible and effective cross features from side information. The results are input into the embedded model based on attention mechanism to capture the invisible interaction among user IDs, item IDs and cross features. The captured interactions are used to predict the match score between the user and the recommended item. Cross-feature attention score is used to generate different recommendation reasons for different user-items.Experimental results show that the proposed algorithm can guarantee the quality of recommendation. The transparency and readability of the recommendation process has been improved by providing reference reasons. This method can help users better understand the recommendation behavior of the system and has certain enlightenment to help the recommender system become more personalized and intelligent.

DARES: An Asynchronous Distributed Recommender System Using Deep Reinforcement Learning

Traditional Recommender Systems (RS) use central servers to collect user data, compute user profiles and train global recommendation models. Central computation of RS models has great results in performance because the models are trained using all the available information and the full user profiles. However, centralised RS require users to share their whole interaction history with the central server and in general are not scalable as the number of users and interactions increases. Central RSs also have a central point of attack with respect to user privacy, because all user profiles and interactions are stored centrally. In this work we propose DARES, an distributed recommender system algorithm that uses reinforcement learning and is based on the asynchronous advantage actor-critic model (A3C). DARES is developed combining the approaches of A3C and federated learning (FL) and allows users to keep their data locally on their own devices. The system architecture consists of (i) a local recommendation model trained locally on the user devices using their interaction and (ii) a global recommendation model that is trained on a central server using the model updates that are computed on the user devices. We evaluate the proposed algorithm using well-known datasets and we compare its performance against well-known state of the art algorithms. We show that although being distributed and asynchronous, it can achieve comparable and in many cases better performance than current state-of-the-art algorithms.

Location-Aware Personalized Traveler Recommender System (LAPTA) Using Collaborative Filtering KNN

Many tourists who travel to explore different cultures and cities worldwide aim to find the best tourist sites, accommodation, and food according to their interests. This objective makes it harder for tourists to decide and plan where to go and what to do. Aside from hiring a local guide, an option which is beyond most travelers’ budgets, the majority of sojourners nowadays use mobile devices to search for or recommend interesting sites on the basis of user reviews. Therefore, this work utilizes the prevalent recommender systems and mobile app technologies to overcome this issue. Accordingly, this study proposes location-aware personalized traveler assistance (LAPTA), a system which integrates user preferences and the global positioning system (GPS) to generate personalized and location-aware recommendations. That integration will enable the enhanced recommendation of the developed scheme relative to those from the traditional recommender systems used in customer ratings. Specifically, LAPTA separates the data obtained from Google locations into name and category tags. After the data separation, the system fetches the keywords from the user’s input according to the user’s past research behavior. The proposed system uses the K-Nearest algorithm to match the name and category tags with the user’s input to generate personalized suggestions. The system also provides suggestions on the basis of nearby popular attractions using the Google point of interest feature to enhance system usability. The experimental results showed that LAPTA could provide more reliable and accurate recommendations compared to the reviewed recommendation applications.

An Industrial Dyeing Recipe Recommendation System for Textile Fabrics Based on Data-Mining and Modular Architecture Design

We report a new fabric dyeing recipe recommendation system which is based on mining industrial dyeing manufacturing data and a system design with modular architecture. Unlike traditional dyeing recipe recommendation systems, our method does not rely on labor-intensive calibration works between dye concentrations and the color. Also, the system is generally designed for different dyeing tasks. We describe the framework of our method and discuss strategies that are used for building the system. The system is built in the form of modular architecture which is made up of multiple gradient boosting regression tree models (GBRT). Each GBRT has been trained for predicting dye concentrations of a dye combination set (DCS) for a fabric type. Methods for model training and typical model performance are reported in the paper as well.

Fairness Embedded Adaptive Recommender System: A Conceptual Framework

In the current fast paced and constantly changing environment, companies should ensure that their way of interacting with user is both relevant and highly adaptive. In order to stay competitive, companies should invest in state-of-the-art technologies that optimize the relationship with the user using increasingly available data. The most popular applications used to develop user relationship are Recommender Systems. The vast majority of the traditional recommender system considers recommendation as a static procedure and focus on a specific type of recommendation, being not very agile in adapting to new situations. Also, when implementing a Recommender System there is the need to ensure fairness in the way decisions are made upon customer data. In this paper, it is proposed a novel Reinforcement Learning-based recommender system that is highly adaptive to changes in customer behavior and focuses on ensuring both producer and consumer fairness, Fairness Embedded Adaptive Recommender System (FEARS). The approach overcomes Reinforcement Learning’s main drawback in recommendation area by using a small, but meaningful action space. Also, there are presented two fairness metrics, their calculation and adaptation for usage with Reinforcement Learning, this way ensuring that the system gets to the optimal trade-off between personalization and fairness.

Content boosted hybrid filtering for solving pessimistic user problem in recommendation systems

The main goal of a recommendation system is to recommend items of interest to users by analyzing their historical data. Content-based and collaborative filtering are the traditional recommendation strategies, each with its own strengths and weaknesses. Some of their weaknesses can be overcome by combining the two strategies. The resulting hybrid system performs qualitatively better than the traditional recommendation systems. However, historical data of some users may consist largely of only likes or only dislikes. Those users are termed as optimistic or pessimistic users respectively. On an average there are around 10 to 20% of pessimistic users present in a given dataset. For pessimistic users, whose profiles have mostly dislikes and very few likes, content-based filtering can hardly recommend any items of interest. In content-based filtering technique pessimistic users get poor recommendations of either uninteresting movies or no recommendations at all. This can be alleviated by boosting the content profiles of pessimistic users using the top-n recommendations of collaborative filtering. This content boosted hybrid filtering system provides a novel list of recommendations even for pessimistic users, with predictive accuracy better than that of a traditional content-based filtering system.

Personalized recommendation by matrix co-factorization with multiple implicit feedback on pairwise comparison

Recommendation systems have been tremendously important to assist users to find relevant items. With the information-overloaded problem, it becomes crucial to understand users’ behavior by learning their preferences during the interaction to construct a profile for exploitation in selecting relevant items. Relevant feedback to capture the users’ behavior may not only explicitly exist but also implicitly available. In the real world, it is common that explicit feedback may be unavailable, and the recommender systems rely only on implicit feedback. When only implicit feedback exists, there are interpretability issues on performing recommender systems. In addition, multiple implicit feedback may cause diverse interpretability due to different characteristics and distributions. This study aims to propose a decomposition approach by incorporating joint information rating to improve recommender systems. In prior to the development of the decomposition approach, we develop a framework to explore the proper rating transformation on multiple implicit feedback. The best rating transformation approach is evaluated using the traditional recommender systems and is used as the input for the joint information rating in the decomposition approach using matrix co-factorization. The proposed matrix co-factorization incorporates multiple implicit feedbacks (i.e., frequency and duration). The result proves that incorporating multiple implicit feedbacks with matrix co-factorization improves the recommendation quality.

Applying Internet information technology combined with deep learning to tourism collaborative recommendation system.

Recently, more personalized travel methods have emerged in the tourism industry, such as individual travel and self-guided travel. The service models of traditional tourism limit the diversity of service options and cannot fully meet the individual needs of tourists anymore. The aim is to integrate sparse tourism information on the Internet, thereby providing more convenient, faster, and more personalized tourism services. Based on the shortcomings of the traditional tourism recommendation system, a deep learning-based classification processing method of tourism product information is proposed. This method uses word embedding in the data preprocessing stage. The Convolutional Neural Network (CNN) is used to process review information of users and tourism service items. The Deep Neural Network (DNN) is used to process the necessary information of users and tourism service items. Also, factorization machine technology is used to learn the interaction between the extracted features to improve the prediction model. The results show that the proposed model can maintain an excellent precision of 64.2% when generating personalized recommendation lists for users. The sensitivity and accuracy of the recommendation list are better than other algorithms. By adding DNN, the word embedding method, and the factorization machine model, the precision is improved by 30%, 33.3%, and 40%, respectively. The model accuracy is the highest with 40 hidden factors, 100 convolutions, and a 100+50 combination hidden layer. Compared with traditional methods, the proposed algorithm can provide users with personalized travel products more accurately in personalized travel recommendations. The results have enriched and developed the theory of tourism service supply chain, providing a reference for constructing a personalized tourism service system.

Recommending Tasks in Online Judges using Autoencoder Neural Networks

Programming contests such as International Olympiads in Informatics (IOI) and ACM International Collegiate Programming Contest (ICPC) are becoming increasingly popular in recent years. To train for these contests, there are several Online Judges available, in which users can test their skills against a usually large set of programming tasks. In the literature, so far few papers have addressed the problem of recommending tasks in online judges. Most notably, as opposed with traditional Recommender Systems, since the learners improve their skills as they solve more problems, there is an intrinsic dynamic dimension that has to be considered: when recommending movies or books, it is likely that the preferences of the users are more or less stable, whilst in recommending tasks this does not hold true. In order to help the learners, it is crucial to recommend them tasks that are challenging but not unsolvable compared with their current set of skills. In this paper we present a Recommender System (RS) for Online Judges based on an Autoencoder (Artificial) Neural Network (ANN). We also discuss the results of an experimental evaluation of our approach in both the scenarios in which we consider, or not, the intrinsic dynamic dimension of the problem. The ANNs are trained with the dataset of all the submissions in the Italian National Online Judge, used to train students for the Italian Olympiads in Informatics.

ConvSVD++: A Hybrid Deep CF Recommender Model using Convolutional Neural Network

Recommender systems are powerful systems that give added value to business and corporation. They are relatively recent technology and they will only keep improving in the future. The most widely used algorithms for recommender systems are categorized into the traditional recommender and deep-based recommender system. The traditional recommendation algorithm suffers from sparse data that significantly degrades recommendation accuracy. The hybrid approaches are attempts to tackle recommendation challenges. This paper addresses the integration of deep learning into traditional recommendation approaches especially, Collaborative Filtering (CF) algorithms to get a significant accurate prediction. It proposes a hybrid deep CF recommender model called ConvSVD++ that tightly integrates Convolution Neural Network (CNN) and Singular Value Decomposition (SVD++). The proposed model incorporates items’ content, implicit user’s feedback along with explicit item-user interaction to enhance prediction accuracy and handle sparsity problem. The proposed model is evaluated and all baseline models based on Movielens- 1M datasets. The results are evaluated using Root Mean Squared Error (RMSE) metric and it is concluded that the proposed model ConvSVD++ outperforms the baselines models. Accordingly, it is concluded that integrating CNN with SVD++ algorithm improves rating prediction accuracy.