In the last 16 years, more than 200 research articles were published about research-paper recommender systems. We reviewed these articles and present some descriptive statistics in this paper, as well as a discussion about the major advancements and shortcomings and an overview of the most common recommendation concepts and approaches. We found that more than half of the recommendation approaches applied content-based filtering (55 %). Collaborative filtering was applied by only 18 % of the reviewed approaches, and graph-based recommendations by 16 %. Other recommendation concepts included stereotyping, item-centric recommendations, and hybrid recommendations. The content-based filtering approaches mainly utilized papers that the users had authored, tagged, browsed, or downloaded. TF-IDF was the most frequently applied weighting scheme. In addition to simple terms, n-grams, topics, and citations were utilized to model users' information needs. Our review revealed some shortcomings of the current research. First, it remains unclear which recommendation concepts and approaches are the most promising. For instance, researchers reported different results on the performance of content-based and collaborative filtering. Sometimes content-based filtering performed better than collaborative filtering and sometimes it performed worse. We identified three potential reasons for the ambiguity of the results. (A) Several evaluations had limitations. They were based on strongly pruned datasets, few participants in user studies, or did not use appropriate baselines. (B) Some authors provided little information about their algorithms, which makes it difficult to re-implement the approaches. Consequently, researchers use different implementations of the same recommendations approaches, which might lead to variations in the results. (C) We speculated that minor variations in datasets, algorithms, or user populations inevitably lead to strong variations in the performance of the approaches. Hence, finding the most promising approaches is a challenge. As a second limitation, we noted that many authors neglected to take into account factors other than accuracy, for example overall user satisfaction. In addition, most approaches (81 %) neglected the user-modeling process and did not infer information automatically but let users provide keywords, text snippets, or a single paper as input. Information on runtime was provided for 10 % of the approaches. Finally, few research papers had an impact on research-paper recommender systems in practice. We also identified a lack of authority and long-term research interest in the field: 73 % of the authors published no more than one paper on research-paper recommender systems, and there was little cooperation among different co-author groups. We concluded that several actions could improve the research landscape: developing a common evaluation framework, agreement on the information to include in research papers, a stronger focus on non-accuracy aspects and user modeling, a platform for researchers to exchange information, and an open-source framework that bundles the available recommendation approaches.

/pdf/research-paper-recommender-systems-a-literature-survey-3o8izjcp9m.pdf

Research-paper recommender systems: a literature survey

Recommender systems use algorithms to provide users with product or service recommendations. Recently, these systems have been using machine learning algorithms from the field of artificial intelligence. However, choosing a suitable machine learning algorithm for a recommender system is difficult because of the number of algorithms described in the literature. Researchers and practitioners developing recommender systems are left with little information about the current approaches in algorithm usage. Moreover, the development of recommender systems using machine learning algorithms often faces problems and raises questions that must be resolved. This paper presents a systematic review of the literature that analyzes the use of machine learning algorithms in recommender systems and identifies new research opportunities. The goals of this study are to (i) identify trends in the use or research of machine learning algorithms in recommender systems; (ii) identify open questions in the use or research of machine learning algorithms; and (iii) assist new researchers to position new research activity in this domain appropriately. The results of this study identify existing classes of recommender systems, characterize adopted machine learning approaches, discuss the use of big data technologies, identify types of machine learning algorithms and their application domains, and analyzes both main and alternative performance metrics.

The use of machine learning algorithms in recommender systems: A systematic review

Recommender systems use algorithms to provide users with product or service recommendations. Recently, these systems have been using machine learning algorithms from the field of artificial intelligence. However, choosing a suitable machine learning algorithm for a recommender system is difficult because of the number of algorithms described in the literature. Researchers and practitioners developing recommender systems are left with little information about the current approaches in algorithm usage. Moreover, the development of a recommender system using a machine learning algorithm often has problems and open questions that must be evaluated, so software engineers know where to focus research efforts. This paper presents a systematic review of the literature that analyzes the use of machine learning algorithms in recommender systems and identifies research opportunities for software engineering research. The study concludes that Bayesian and decision tree algorithms are widely used in recommender systems because of their relative simplicity, and that requirement and design phases of recommender system development appear to offer opportunities for further research.

The Use of Machine Learning Algorithms in Recommender Systems: A Systematic Review

This chapter aims to provide an overview of the class of multi-criteria recommender systems, i.e., the category of recommender systems that use multi-criteria preference ratings. Traditionally, the vast majority of recommender systems literature has focused on providing recommendations by modelling a user’s utility (or preference) for an item as a single preference rating. However, where possible, capturing richer user preferences along several dimensions—for example, capturing not only the user’s overall preference for a given movie but also her preferences for specific movie aspects (such as acting, story, or visual effects)—can provide opportunities for further improvements in recommendation quality. As a result, a number of recommendation techniques that attempt to take advantage of such multi-criteria preference information have been developed in recent years. A review of current algorithms that use multi-criteria ratings for calculating predictions and generating recommendations is provided. The chapter concludes with a discussion on open issues and future challenges for the class of multi-criteria rating recommenders.

Multi-Criteria Recommender Systems

Cloud computing is becoming popular. Building high-quality cloud applications is a critical research problem. QoS rankings provide valuable information for making optimal cloud service selection from a set of functionally equivalent service candidates. To obtain QoS values, real-world invocations on the service candidates are usually required. To avoid the time-consuming and expensive real-world service invocations, this paper proposes a QoS ranking prediction framework for cloud services by taking advantage of the past service usage experiences of other consumers. Our proposed framework requires no additional invocations of cloud services when making QoS ranking prediction. Two personalized QoS ranking prediction approaches are proposed to predict the QoS rankings directly. Comprehensive experiments are conducted employing real-world QoS data, including 300 distributed users and 500 real-world web services all over the world. The experimental results show that our approaches outperform other competing approaches.

QoS Ranking Prediction for Cloud Services

A recommender system is useful for a digital library to suggest the books that are likely preferred by a user. Most recommender systems using collaborative filtering approaches leverage the explicit user ratings to make personalized recommendations. However, many users are reluctant to provide explicit ratings, so ratings-oriented recommender systems do not work well. In this paper, we present a recommender system for CADAL digital library, namely CARES, which makes recommendations using a ranking-oriented collaborative filtering approach based on users' access logs, avoiding the problem of the lack of user ratings. Our approach employs mean AP correlation coefficients for computing similarities among users' implicit preference models and a random walk based algorithm for generating a book ranking personalized for the individual. Experimental results on real access logs from the CADAL web site show the effectiveness of our system and the impact of different values of parameters on the recommendation performance.

CARES: a ranking-oriented CADAL recommender system

Nowadays some recommender system researchers have already been engaging multi-criteria that model possible attributes of the item to generate the improved recommendations. However, the statistical machine learning methods successful in the single-rating recommender system have not been investigated in the context of multi-criteria ratings. In this paper, we propose two types of multi-criteria probabilistic latent semantic analysis algorithms extended from the single-rating version. First, the mixture of multi-variate Gaussian distribution is assumed to be the underlying distribution of multi-criteria ratings of each user. Second, we further assume the mixture of the linear Gaussian regression model as the underlying distribution of multi-criteria ratings of each user, inspired by the Bayesian network and linear regression. The experiment results on the Yahoo!Movies ratings data set show that the full multi-variate Gaussian model and the linear Gaussian regression model achieve a stable performance gain over other tested methods.

Applying probabilistic latent semantic analysis to multi-criteria recommender system

The invention discloses an automatic book keyword generation method The relation between a book and an author thereof is determined, book reviews are collected from the Internet and dismantled into words, which are tagged with parts of speech, and nouns are kept as alternate keywords; the number of documents with the alternate keywords appeared in Wikipedia is retrieved by Google in order to preliminarily assesses the importance, and modal particles and wrongly written words are filtered; a four-step map is constructed with the author, the book, the review and the alternate keyword as four types of nodes and the relations between the author, the book, the review and the alternate keyword as sides; and based on the four-step map, an absorbing-state random walk algorithm with the book as a center is applied to select the most important alternate keyword as a book keyword The invention takes the subject consistency, content correlation and reader feedback of the books of the same author into consideration, so the accuracy rate of the keyword is high; the book reviews of readers from different perspectives are sufficiently covered, and the keyword is diversified; data comes from the Internet, the number of book reviews is enormous, coverage is wide, and the coverage of the method is wide

Automatic book keyword generation method

Book reviews contributed by readers in social sites contain valuable information on books' content, style and merit, many informative words in which can be used to enrich metadata of books in China-Us Million Book Digital Library. In this paper, we present a system for review-oriented metadata enrichment and propose an Book-Centric Diverse Random Walk algorithm on a four-partite graph containing three kinds of relations among authors, books, reviews and words, in order to produce highly relevant as well as diverse keywords for a book. Experimental results of a user study show that our approach significantly outperforms other methods in terms of relevance and diversity. The metadata generated by our approach also has a large overlap with popular social tags and brief introductions from DouBan for books in the coverage experiments.

Liang Zhang

Papers

CARES: a ranking-oriented CADAL recommender system

Applying probabilistic latent semantic analysis to multi-criteria recommender system

Automatic book keyword generation method

Review-oriented metadata enrichment: a case study