Search results

Online news websites provide huge amounts of timely news, bringing the challenge of recommending personalized news articles. Generative adversarial networks (GAN) based on collaborative filtering (CFGAN) can achieve effective recommendation quality. However, CFGAN ignores item contents, which contain more latent preference features than just user ratings. It is important to consider both ratings and item contents in making preference predictions. This study aims to improve news recommendation by proposing a GAN-based news recommendation model considering both ratings (implicit feedback) and the latent features of news content.

The collaborative topic modeling (CTM) can improve user preference prediction by combining matrix factorization (MF) with latent topics of item content derived from latent topic modeling. This study proposes a novel hybrid news recommendation model, Hybrid-CFGAN, which modifies the architecture of the CFGAN model with enhanced preference learning from the CTM. The proposed Hybrid-CFGAN model contains parallel neural networks – original rating-based preference learning and CTM-based preference learning, which consider both ratings and news content with user preferences derived from the CTM model. A tunable parameter is used to adjust the weights of the two preference learnings, while concatenating the preference outputs of the two parallel neural networks.

This study uses the dataset collected from an online news website, NiusNews, to conduct an experimental evaluation. The results show that the proposed Hybrid-CFGAN model can achieve better performance than the state-of-the-art GAN-based recommendation methods. The proposed novel Hybrid-CFGAN model can enhance existing GAN-based recommendation and increase the performance of preference predictions on textual content such as news articles.

As the existing CFGAN model does not consider content information and solely relies on history logs, it may not be effective in recommending news articles. Our proposed Hybrid-CFGAN model modified the architecture of the CFGAN generator by adding a parallel neural network to gain the relevant information from news content and user preferences derived from the CTM model. The novel idea of adjusting the preference learning from two parallel neural networks – original rating-based preference learning and CTM-based preference learning – contributes to improve the recommendation quality of the proposed model by considering both ratings and latent preferences derived from item contents. The proposed novel recommendation model can improve news recommendation, thereby increasing the commercial value of news media platforms.

The explosion of data due to the sophistication of information and communication technology makes it simple for prospective tourists to learn about previous hotel guests' experiences. They prioritize the rating score when selecting a hotel. However, rating scores are less reliable for suggesting a personalized preference for each aspect, especially when they are in a limited number. This study aims to recommend ratings and personalized preference hotels using cross-domain and aspect-based features.

We propose an aspect-based cross-domain personalized recommendation (AsCDPR), a novel framework for rating prediction and personalized customer preference recommendations. We incorporate a cross-domain personalized approach and aspect-based features of items from the review text. We extracted aspect-based feature vectors from two domains using bidirectional long short-term memory and then mapped them by a multilayer perceptron (MLP). The cross-domain recommendation module trains MLP to analyze sentiment and predict item ratings and the polarities of the aspect based on user preferences.

Expanded by its synonyms, aspect-based features significantly improve the performance of sentiment analysis on accuracy and the F1-score matrix. With relatively low mean absolute error and root mean square error values, AsCDPR outperforms matrix factorization, collaborative matrix factorization, EMCDPR and Personalized transfer of user preferences for cross-domain recommendation. These values are 1.3657 and 1.6682, respectively.

This study assists users in recommending hotels based on their priority preferences. Users do not need to read other people's reviews to capture the key aspects of items. This model could enhance system reliability in the hospitality industry by providing personalized recommendations.

This study introduces a new approach that embeds aspect-based features of items in a cross-domain personalized recommendation. AsCDPR predicts ratings and provides recommendations based on priority aspects of each user's preferences.

Although user stickiness has been studied for several years in the field of live e-commerce, little attention has been paid to the effects of streamer attributes on user stickiness in this field. Rooted in the stimulus-organism-response (S-O-R) theory, this study investigated how streamer attributes influence user stickiness.

The authors obtained 496 valid samples from Chinese live e-commerce users and explored the formation of user stickiness using partial least squares-structural equation modeling (PLS-SEM). Artificial neural network (ANN) was used to capture linear and non-linear relationships and analyze the normalized importance ranking of significant variables, supplementing the PLS-SEM results.

The authors found that attractiveness and similarity positively impacted parasocial interaction (PSI). Expertise and trustworthiness positively impacted perceived information quality. Moreover, streamer-brand preference mediated the relationship between PSI and user stickiness, as well as the relationship between perceived information quality and user stickiness. Compared to PLS-SEM, the predictive ability of ANN was more robust. Further, the results of PLS-SEM and ANN both showed that attractiveness was the strongest predictor of user stickiness.

This study explained how streamer attributes affect user stickiness and provided a reference value for future research on user behavior in live e-commerce. The exploration of the linear and non-linear relationships between variables based on ANN supplements existing research. Moreover, the results of this study have implications for practitioners on how to improve user stickiness and contribute to the development of the livestreaming industry.

In this “Info-plosion” era, recommendation systems (or recommenders) play a significant role in finding interesting items in the surge of online digital activity and e-commerce. The purpose of this paper is to model users' preference evolution to recommend potential items which users may be interested in.

A novel recommendation system, namely evolution-learning recommendation (ELR), is developed to precisely predict user interest for making recommendations. Differing from prior related methods, the authors integrate the matrix factorization (MF) and recurrent neural network (RNN) to effectively describe the variation of user preferences over time.

A novel cumulative factorization technique is proposed to efficiently decompose a rating matrix for discovering latent user preferences. Compared to traditional MF-based methods, the cumulative MF could reduce the utilization of computation resources. Furthermore, the authors depict the significance of long- and short-term effects in the memory cell of RNN for evolution patterns. With the context awareness, a learning model, V-LSTM, is developed to dynamically capture the evolution pattern of user interests. By using a well-trained learning model, the authors predict future user preferences and recommend related items.

Based on the relations among users and items for recommendation, the authors introduce a novel concept, virtual communication, to effectively learn and estimate the correlation among users and items. By incorporating the discovered latent features of users and items in an evolved manner, the proposed ELR model could promote “right” things to “right” users at the “right” time. In addition, several extensive experiments are performed on real datasets and are discussed. Empirical results show that ELR significantly outperforms the prior recommendation models. The proposed ELR exhibits great generalization and robustness in real datasets, including e-commerce, industrial retail and streaming service, with all discussed metrics.

This systematic literature review aims to elaborate on the research progress and features of information source preferences to help other researchers attain a more comprehensive understanding of the field.

Following a systematic review protocol, 139 research articles from 11 academic databases were analyzed.

Overall, five separate results were obtained: first, information source horizon theory is the main theoretical foundation of information source preferences research, while other theories have been applied less. Second, information source preference research has strong context sensitivity and involves health, work, consumption, learning, survival and development and emergencies. Third, preference criteria can be summarized into three categories: information characteristics, user characteristics, needs characteristics and corresponding specific criteria. Fourth, information source preferences are influenced by both internal and external factors, including five specific aspects, namely demographics, the user's cognition, the user's affection, capital and contextual factors. Fifth, this field is dominated by quantitative methods and an information horizon mapping method could be applied more.

This study is the first to reveal the general picture of information source preferences. It also elaborates on the characteristics of this field and presents potential development directions.

This paper aims to address the following issues: (1) most existing methods are based on recurrent network, which is time-consuming to train long sequences due to not allowing for full parallelism; (2) personalized preference generally are not considered reasonably; (3) existing methods rarely systematically studied how to efficiently utilize various auxiliary information (e.g. user ID and time stamp) in trajectory data and the spatiotemporal relations among nonconsecutive locations.

The authors propose a novel self-attention network–based model named SanMove to predict the next location via capturing the long- and short-term mobility patterns of users. Specifically, SanMove uses a self-attention module to capture each user's long-term preference, which can represent her personalized location preference. Meanwhile, the authors use a spatial-temporal guided noninvasive self-attention (STNOVA) module to exploit auxiliary information in the trajectory data to learn the user's short-term preference.

The authors evaluate SanMove on two real-world datasets. The experimental results demonstrate that SanMove is not only faster than the state-of-the-art recurrent neural network (RNN) based predict model but also outperforms the baselines for next location prediction.

The authors propose a self-attention-based sequential model named SanMove to predict the user's trajectory, which comprised long-term and short-term preference learning modules. SanMove allows full parallel processing of trajectories to improve processing efficiency. They propose an STNOVA module to capture the sequential transitions of current trajectories. Moreover, the self-attention module is used to process historical trajectory sequences in order to capture the personalized location preference of each user. The authors conduct extensive experiments on two check-in datasets. The experimental results demonstrate that the model has a fast training speed and excellent performance compared with the existing RNN-based methods for next location prediction.

Users' voluntary forwarding behavior opens a new avenue for companies to promote their brands and products on social networking sites (SNS). However, research on voluntary information disseminators is limited. This paper aims to bring an in-depth understanding of voluntary disseminators by answering the following questions: (1) What is the underlying mechanism by which some users are more enthusiastic to voluntarily forward content of interest? (2) How to identify them? We propose a theoretical model based on the Elaboration-Likelihood Model (ELM) and examine three types of factors that moderate the effect of preference matching on individual forwarding behavior, including personal characteristics, tweet characteristics and sender–receiver relationships.

Via Twitter API, we randomly crawled 1967 Twitter users' data to validate the conceptual framework. Each user’s original tweets and retweeted tweets, profile data such as the number of followers and followees and verification status were obtained. The final corpus contains 163,554 data points composed of 1,634 valid twitterers' retweeting behavior. Tweets produced by these core users' followees were also crawled. These data points constitute an unbalanced panel data and we employ different models — fixed-effects, random-effects and pooled logit models — to test the moderation effects. The robustness test shows consistency among these different models.

Preference matching significantly affects users' forwarding behavior, implying that SNS users are more likely to share contents that align with their preferences. In addition, we find that popular users with lots of followers, heavy SNS users who author tweets or forward other-sourced tweets more frequently and users who tend to produce longer original contents are more enthusiastic to disseminate contents of interest. Furthermore, interaction strength has a positive moderating effect on the relationship between preference matching and individuals' forwarding decisions, suggesting that users are more likely to disseminate content of interest when it comes from strong ties. However, the moderating effect of perceived affinity is significantly negative, indicating that an online community of individuals with many common friends is not an ideal place to engage individuals in sharing information.

This work brings about a deep understanding of users' voluntary forwarding behavior of content of interest. To the best of our knowledge, the current study is the first to examine (1) the underlying mechanism by which some users are more likely to voluntarily forward content of interest; and (2) how to identify these potential voluntary disseminators. By extending the ELM, we examine the moderating effect of tweet characteristics, sender–receiver relationships as well as personal characteristics. Our research findings provide practical guidelines for enterprises and government institutions to choose voluntary endorsers when trying to engage individuals in information dissemination on SNS.

With the rapid development of social media, the occurrence and evolution of emergency events are often accompanied by massive users' expressions. The fine-grained analysis on users' expressions can provide accurate and reliable information for event processing. Hence, 2,003,814 expressions on a major malignant emergency event were mined from multiple dimensions in this paper.

This paper conducted finer-grained analysis on users' online expressions in an emergency event. Specifically, the authors firstly selected a major emergency event as the research object and collected the event-related user expressions that lasted nearly two years to describe the dynamic evolution trend of the event. Then, users' expression preferences were identified by detecting anomic expressions, classifying sentiment tendencies and extracting topics in expressions. Finally, the authors measured the explicit and implicit impacts of different expression preferences and obtained relations between the differential expression preferences.

Experimental results showed that users have both short- and long-term attention to emergency events. Their enthusiasm for discussing the event will be quickly dispelled and easily aroused. Meanwhile, most users prefer to make rational and normative expressions of events, and the expression topics are diversified. In addition, compared with anomic negative expressions, anomic expressions in positive sentiments are more common. In conclusion, the integration of multi-dimensional analysis results of users' expression preferences (including discussion heat, preference impacts and preference relations) is an effective means to support emergency event processing.

To the best of the authors' knowledge, it is the first research to conduct in-depth and fine-grained analysis of user expression in emergencies, so as to get in-detail and multi-dimensional characteristics of users' online expressions for supporting event processing.

This study aims to fill the lack of quantitative studies of user preferences in quasi-public spaces to observe the use of quasi-public spaces by questioning the contemporary needs of urban communities and to develop design strategies accordingly.

Within the scope of this study, public space design elements affecting users' preferences in the quasi-public spaces of the Podium Park shopping center in Bursa, Turkey were evaluated. By considering the spatial characteristics of the study area, 4 main and 15 subcriteria were determined and utilized by analytic hierarchy process (AHP). These criteria were evaluated by experts and locals with a participatory approach.

According to the obtained results, “events” (S2), “sun/shade” (C2), “safety” (P3) and “planting” (U4) subcriteria were determined as the vital elements for quasi-public spaces.

Although the concept of quasi-public space has been discussed for nearly 30 years, it has been observed that there are no quantitative studies to determine the criteria of user preferences in these open spaces in the literature. This study is the first quantitative research for user preferences in quasi-public spaces and there is no previous study on this subject and study area in Turkey.

The purpose of this study is to investigate how book group recommendations can be used as a meaningful way to suggest suitable books to users, given the increasing number of individuals engaging in sharing and discussing books on the web.

The authors propose reviews fine-grained classification (CFGC) and its related models such as CFGC1 for book group recommendation. These models can categorize reviews successively by function and role. Constructing the BERT-BiLSTM model to classify the reviews by function. The frequency characteristics of the reviews are mined by word frequency analysis, and the relationship between reviews and total book score is mined by correlation analysis. Then, the reviews are classified into three roles: celebrity, general and passerby. Finally, the authors can form user groups, mine group features and combine group features with book fine-grained ratings to make book group recommendations.

Overall, the best recommendations are made by Synopsis comments, with the accuracy, recall, F-value and Hellinger distance of 52.9%, 60.0%, 56.3% and 0.163, respectively. The F1 index of the recommendations based on the author and the writing comments is improved by 2.5% and 0.4%, respectively, compared to the Synopsis comments.

Previous studies on book recommendation often recommend relevant books for users by mining the similarity between books, so the set of book recommendations recommended to users, especially to groups, always focuses on the few types. The proposed method can effectively ensure the diversity of recommendations by mining users’ tendency to different review attributes of books and recommending books for the groups. In addition, this study also investigates which types of reviews should be used to make book recommendations when targeting groups with specific tendencies.