User Behavior Information Research Articles

Wearable Biosensor Smart Glasses Based on Augmented Reality and Eye Tracking.

With the rapid development of wearable biosensor technology, the combination of head-mounted displays and augmented reality (AR) technology has shown great potential for health monitoring and biomedical diagnosis applications. However, further optimizing its performance and improving data interaction accuracy remain crucial issues that must be addressed. In this study, we develop smart glasses based on augmented reality and eye tracking technology. Through real-time information interaction with the server, the smart glasses realize accurate scene perception and analysis of the user's intention and combine with mixed-reality display technology to provide dynamic and real-time intelligent interaction services. A multi-level hardware architecture and optimized data processing process are adopted during the research process to enhance the system's real-time accuracy. Meanwhile, combining the deep learning method with the geometric model significantly improves the system's ability to perceive user behavior and environmental information in complex environments. The experimental results show that when the distance between the subject and the display is 1 m, the eye tracking accuracy of the smart glasses can reach 1.0° with an error of no more than ±0.1°. This study demonstrates that the effective integration of AR and eye tracking technology dramatically improves the functional performance of smart glasses in multiple scenarios. Future research will further optimize smart glasses' algorithms and hardware performance, enhance their application potential in daily health monitoring and medical diagnosis, and provide more possibilities for the innovative development of wearable devices in medical and health management.

Empowering Library Services through User‐Centric Analysis of Co‐loan Patterns

ABSTRACTBy examining loan data from public libraries in South Korea, we seek to understand patterns in user borrowing behaviors and explore thematic connections among borrowed books. The subject headings of 55.5 million book sets borrowed by individual users on the same day were analyzed using ITEM2VEC. We have identified 40 subject heading communities through cosine similarity of each subject vector, and we have labeled each community using a large language model. Two prominent communities were identified: Global Modern Literature and Novels and Children's Literature, Fairy Tales, and Folklore. The latter community was associated with a diverse array of subjects, indicating an expansion in children's reading preferences. The study results will be useful for improving collection development and the relevance of book recommendations, as well as for incorporating user information behavior into traditional library material classification schemes.

Anomalous behavior detection based on optimized graph embedding representation in social networks

Anomalous behaviors in social networks can lead to privacy leaks and the spread of false information. In this paper, we propose an anomalous behavior detection method based on optimized graph embedding representation. Specifically, the user behavior logs are first extracted into a social network user behavior temporal knowledge graph, based on which the graph embedding representation method is used to transform the network topology and temporal information in the user behavior knowledge graph into structural embedding vectors and temporal information embedding vectors, and the hybrid attention mechanism is used to merge the two types of vectors to obtain the final entity embedding to complete the prediction and complementation of the temporal knowledge graph of user behavior. We use graph neural networks, which use the temporal information of user behaviors as a time constraint and capture both user behavioral and semantic information. It converts the two parts of information into vectors for concatenation and linear transformation to obtain a comprehensive representation vector of the whole subgraph, as well as joint deep learning models to evaluate abnormal behavior. Finally, we perform experiments on the Yelp dataset to validate that our method achieves a 9.56% improvement in the F1-score.

Meta graph network recommendation based on multi-behavior encoding

As traditional recommendation systems ignore the hidden information among different user behaviors (such as clicks, add-to-favorites, add-to-cart, and purchases), this often leads to low accuracy in recommendation results. We propose a meta-graph network recommendation system via multi-behavior encoding (MBGR). Firstly, the graph convolutional neural network is used to extract features from various interactive behavior heterogeneous graphs of user-items for behavior heterogeneous modeling. Secondly, matrix decomposition algorithm and meta-knowledge learner are used respectively to process the semantic information of user behavior, and then attention mechanism is used to learn and distinguish the importance of different types of user item interaction behaviors. Finally, meta-knowledge transfer network is used to combine meta-learning paradigm and neural network framework to establish user target behavior recommendation. We conducted comparative experiments comparing MBGR with 7 different baseline models such as NCF and DMF. Extensive experiments on three real datasets (Tmall, Yelp, ML10M) demonstrate that the proposed MBGR method outperforms the baselines. The performance of MBGR is improved by 10.97 % on average with the metric of HR@10 and 10.96 % with the metric of NDCG@10. Under different top-N value evaluation conditions (HR@10, HR@7, NDCG@10, NDCG@7, etc.), the proposed model's performance can also be improved by more than 10 %, which proves the rationality and effectiveness of the proposed MBGR method.

An exploratory study of information re-finding behaviour modes of Chinese college students on social media: video diary analysis from Chinese platforms

ABSTRACT This study aims to explore college students’ information-re-finding behaviour in the context of social media by identifying different behaviour modes to enrich the research on the information behaviour of social media users. Thirty-two college students volunteered to record their re-finding process once it occurred within two weeks by the video diary method. One hundred and forty-six valid events were collected and analysed qualitatively and quantitatively. It was shown that the behaviour process of information re-finding on social media experienced six stages: scene triggering, information demand generating, social media selecting, re-finding, information confirmation and information processing. Forty-eight kinds of information re-finding modes were illustrated and summarised into seven categories according to where and how the information was finally found: following, chat record, keyword, favourites, history, personal dynamics and the like. From further analysis, it was proved that significant differences did exist between varied re-finding behaviour modes in information needs and re-finding time rather than information processing behaviour. The study could expand the traditional information search model into mixed timing by considering recalling and re-finding. Practically, it generates useful practical implications for the design of re-finding functions, information recommendation and memory services of social media.

Multi-Channel Hypergraph Collaborative Filtering with Attribute Inference

In the field of collaborative filtering, attribute information is often integrated to improve recommendations. However, challenges remain unaddressed. Firstly, existing data modeling methods often fall short of appropriately handling attribute information. Secondly, attribute data are often sparse and can potentially impact recommendation performance due to the challenge of incomplete correspondence between the attribute information and the recommendations. To tackle these challenges, we propose a hypergraph collaborative filtering with attribute inference (HCFA) framework, which segregates attribute and user behavior information into distinct channels and leverages hypergraphs to capture high-order correlations among vertices, offering a more natural approach to modeling. Furthermore, we introduce behavior-based attribute confidence (BAC) for assessing the reliability of inferred attributes concerning the corresponding behaviors and update the most credible portions to enhance recommendation quality. Extensive experiments conducted on three public benchmarks demonstrate the superiority of our model. It consistently outperforms other state-of-the-art approaches, with ablation experiments further confirming the effectiveness of our proposed method.

A Job Recommendation Model Based on a Two-Layer Attention Mechanism

In the field of job recruitment, traditional recommendation methods only rely on users’ rating data of positions for information matching. This simple strategy has problems such as low utilization of multi-source heterogeneous data and difficulty in mining relevant information between recruiters and applicants. Therefore, this paper proposes a recurrent neural network model based on a two-layer attention mechanism. The model first improves the entity representation of recruiters and applicants through user behavior, company-related knowledge and other information. The entities and their combinations are then mapped to the vector space using one-hot and TransR methods, and a recurrent neural network with a two-layer attention mechanism is used to obtain their potential interests from the click sequence, and then a recommendation list is generated. The experimental results show that this model achieves better results than previous models.

BRITD: behavior rhythm insider threat detection with time awareness and user adaptation

Researchers usually detect insider threats by analyzing user behavior. The time information of user behavior is an important concern in internal threat detection. Existing works on insider threat detection fail to make full use of the time information, which leads to their poor detection performance. In this paper, we propose a novel behavioral feature extraction scheme: we implicitly encode absolute time information in the behavioral feature sequences and use a feature sequence construction method taking covariance into account to make our scheme adaptive to users. We select Stacked Bidirectional LSTM and Feedforward Neural Network to build a deep learning-based insider threat detection model: Behavior Rhythm Insider Threat Detection (BRITD). BRITD is universally applicable to various insider threat scenarios, and it has good insider threat detection performance: it achieves an AUC of 0.9730 and a precision of 0.8072 with the CMU CERT dataset, which exceeds all baselines.Graphical

User behaviour and hoax information on social media case of Indonesia

The present study examines the influence of user behaviour on the propagation of deceptive information on social media platforms within the Indonesian setting. It specifically focuses on the ramifications of such behaviour on political, health, business competition, and environmental matters. The beginning of the study highlights the gravity of the dissemination of hoaxes on social media platforms, attributing it to the algorithms that prioritise information that is engaging, frequently characterised by provocative or sensationalised headlines. The methodology section discusses how we adopted a mixed-methods approach to examine both quantitative data from social media metrics and qualitative data from user interviews. Additionally, network analysis was used to identify disinformation dissemination across platforms. The findings suggest that hoaxes spread quickly because they grab users' attention better than real information. Our research also shows that these fraudulent schemes can increase social strife, anger, and prejudice against certain groups or people, tarnishing their social standing. Our research shows that in democratic nations like Indonesia, where freedom of expression is important, incorrect information is often used for electoral gain. Our analysis shows that users need better digital literacy to handle this issue. Training people to spot fake news is one way to combat the growing problem of disinformation.

Deep learning methods used in movie recommendation systems

As the amount of internet movie data grows rapidly, traditional movie recommendation systems face increasing challenges. They typically rely on statistical algorithms such as item-based or user-based collaborative filtering. However, these algorithms struggle to handle large-scale data and often fail to capture the complexity and contextual information of user behavior. Therefore, deep learning techniques have been widely applied to movie recommendation systems. This paper reviews movie recommendation algorithms based on traditional statistical models and introduces three main deep learning techniques: Artificial Neural Networks (ANN), Convolutional Neural Networks (CNN), and Recurrent Neural Networks (RNN). ANN can extract features at different levels of users and movies; CNN can capture features of movie posters and movie data to recommend similar movies; RNN can consider user historical behavior and contextual information to better understand user interests and demands. The application of these deep learning techniques can enhance the accuracy and user experience of movie recommendation systems. This paper also demonstrates the advantages and disadvantages of these models and their specific application methods in movie recommendation systems, and points out the direction for further development and improvement of deep learning models in this field.

E-commerce bookstore user alignment model based on multidimensional feature joint representation and implicit behavior compensation

The accurate identification of the same user across different e-commerce platforms is crucial for cross-platform recommendations. Given the complexity of the feature space of heterogeneous e-commerce platforms and the sparsity of explicit user behavior data, we propose a heterogeneous e-commerce network alignment model based on multidimensional joint representation and implicit behavior compensation. First, in response to the complexity of the feature space of heterogeneous e-commerce platforms, the UBN2vec model was constructed. The model generates independent embeddings for multidimensional features of “user attributes-behaviors-neighborhood”, and integrates them using an attention mechanism. Furthermore, through joint optimization of user multidimensional features, low-rank dense user overall embedding vectors are generated, effectively representing the complex features of e-commerce users. Second, to alleviate the sparsity issue of user behavior data on an e-commerce bookstore platform, we propose a method for compensating implicit user behavior data based on item association. This method predicts potential user behavior information by mining the latent correlations between items. It effectively captures user preferences and improves alignment accuracy. Finally, to reduce the computational complexity of the user-matching process, a user interest group block indexing method based on user interest preferences is developed. This method reduces matching times among users with significantly different interests, thereby reducing the computational complexity of the heterogeneous e-commerce bookstore user alignment algorithm. Experiments show that the proposed model can accurately and efficiently identify the same user across e-commerce bookstore platforms.

The Impact of Historical Performance and Managerial Risk-taking Propensity on the Behavior of Choosing Prospector Strategy and Using Strategic Management Accounting Information in Viet Nam Manufacturer

Strategic management accounting (SMA) supports managers in decision-making and activity control to accomplish both short- and long-term objectives. This study, based on the Upper Echelons Theory (UET), investigates how the CFO’s psychological trait—the propensity to take managerial risks, affects the decision to choose prospector strategy and uses SMA information. The main objective of this study is to find out what factors influence the behavior using SMA information. This study also takes into account the internal locus of control “s moderating role effect in some relationships. The CFOs of Vietnamese major manufacturers are investigated. The results have shown that: (1) the lower of the historical performance, the more likely CFOs are to choose the prospector strategy, the more CFOs have managerial risk-taking propensity, the more SMA information usage behavior increases; (2) the more CFOs have managerial risk-taking propensity, the more likely they are to choose the prospector strategy, and the behavior of using SMA information also increases; (3) the CFOs” internal locus of control (ILOC) trait plays a moderator role in increasing their SMA information usage behavior and prospector strategy selection trend. This study adds to accounting expertise by highlighting the significance of SMA information for decision-making, the CFO’s management risk-taking propensity, and helping top managers develop an effective SMA system.

Conceptual framework of recommendation system with hybrid method

Recommendation system relies on information of user preference and user behavior in order to recommend the useful information. The existing recommendation systems still have problems for new users and new items. This research proposes a new hybrid method to develop the conceptual framework of recommendation system that deals with new user and new movie data. The data used consists of a data from MovieLens and the internet movie database (IMDB). This work introduces a hybrid recommendation system which based on a combination of content-based filtering (CBF) and collaborative filtering (CF). Pre-filtering data is performed by finding an optimal number of clusters by calculating the total within cluster sum of square. In order to reduce the complexity of data and increase the relevance of the user-item ratings, the fuzzy c-mean (FCM) is employed. Then the similarity is calculated by using item-based method, the K-nearest neighbors and weight sum of the rating are applied. Finally, to recommend the movies, the research found that for new user data the precision is at 85% and mean absolute eror (MAE) value 2.1011. For new item data, the result of research obtains the precision at 87% and MAE value 2.0031. In conclusion, the new hybrid method developed can recommend movie efficiently.

Research on Consumer Behavior Prediction Based on E-commerce Data Analysis

The user behavior log of e-commerce platform contains a wide range of data, which can be roughly divided into three types: user basic information, merchant information and user behavior information. Behavioral information includes clicks, favorites, shopping carts and purchases. Because of the asymmetry of online shopping information, the virtuality of trading environment, the incomplete perception of goods and the unsynchronization of trading process, consumers are faced with many uncertainties in the process of online shopping. According to the specific research problems, this paper applies the real consumer behavior data in the e-commerce data platform to build a model to predict consumer behavior. The consumer behavior sequence and other characteristics are analyzed by two machine learning methods respectively. First, the consumer behavior sequence based on time sequence is analyzed. Add the obtained probability value as a new feature to other feature sets as the input of NB (Naive Bayes) model, and train the model to get the probability of whether consumers will buy, and then get the final judgment result of the model. The results show that the fusion model has a balanced performance for consumer behavior sequences with different lengths, and its accuracy can be kept at around 0.9. The research on the problems raised in this paper can also be extended to other fields.

Exploiting High-Order Interaction Relations to Explore User Intent (Student Abstract)

This paper studies the problem of exploring the user intent for session-based recommendations. Its challenges come from the uncertainty of user behavior and limited information. However, current endeavors cannot fully explore the mutual interactions among sessions and do not explicitly model the complex high-order relations among items. To circumvent these critical issues, we innovatively propose a HyperGraph Convolutional Contrastive framework (termed HGCC) that consists of two crucial tasks: 1) The session-based recommendation (SBR task) that aims to capture the beyond pair-wise relationships between items and sessions. 2) The self-supervised learning (SSL task) acted as the auxiliary task to boost the former task. By jointly optimizing the two tasks, the performance of the recommendation task achieves decent gains. Experiments on multiple real-world datasets demonstrate the superiority of the proposed approach over the state-of-the-art methods.

The Early Impact of GDPR Compliance on Display Advertising: The Case of an Ad Publisher

The European Union's General Data Protection Regulation (GDPR), with its explicit consent requirement, may restrict the use of personal data and shake the foundations of online advertising. The ad industry has predicted drastic loss of revenue from GDPR compliance and has been seeking alternative ways of targeting. Taking advantage of an event created by an ad publisher's request for explicit consent from users with European Union IP addresses, the authors find that for a publisher that uses a pay-per-click model, has the capacity to leverage both user behavior and web page content information for advertising, and observes high consent rates, GDPR compliance leads to modest negative effects on ad performance, bid prices, and ad revenue. The changes in ad metrics can be explained by temporal variations in consent rates. The impact is most pronounced for travel and financial services advertisers and least pronounced for retail and consumer packaged goods advertisers. The authors further find that web page context can compensate for the loss of access to users’ personal data, as the GDPR's negative impact is less pronounced when ads are posted on web pages presenting relevant content. The results suggest that publishers and advertisers should leverage targeting based on web page content after the GDPR’s rollout.

MUD enabled deep learning framework for anomaly detection in IoT integrated smart building

Nowadays, many Internet of Things (IoT) devices of different types are used in creating smart applications like smart cities, smart industries, smart environments, and the applications of industry-4.0. IoT devices are used for different purposes, such as security, remote monitoring, resource allocation, threats, ecosystems, and vulnerabilities. This paper proposed a deep learning algorithm-based solution to tighten the security level in the IoT-Smart environment network. The Intrusion Detection System (IDS) considered in this paper is Network IDS, which investigates the manufacturer usage description, digital twins, and deep learning-based user behavior information. IoT devices' communication and the users in smart buildings are automatically connected in the Intelligent Communication system. Since many devices and users are interconnected in smart buildings, the probability of cyber-attack is high. Thus, better security is needed in smart buildings and smart environments. It should focus on securing IoT devices, users, and their communication. Hence, this paper developed a deep learning-based anomaly detection framework to dynamically monitor the issues and problems with MUD profiles and detect the anomaly behavior. The Manufacturer Usage Description (MUD) profiles, dynamic user behavior, IoT devices' traffic data the pattern of abnormal/anomaly traffic at the device level is predicted while traffic occurs. The MUD-ML-based model is implemented in Python software, verifying the results.

Business model innovation in the food sector: towards a dimensional framework for analyzing the value proposition of digital platforms for food information services

PurposeDigital platforms for the provision of food information-based services (FISs) represent a consolidated business with increasing revenue streams for entrepreneurs. Such platforms have transformed and clarified the nature of uncertainty and ambiguity inherent in the traditional food sector entrepreneurial processes. Anyway, a clear understanding of the value of digital platforms for FISs is not yet consolidated in the literature. With this paper, the authors try to fill this gap through a critical literature review of scientific research that combines knowledge on food consumer's behavior and user's knowledge behavior.Design/methodology/approachThe authors carried out a critical literature review of scientific research combining knowledge of food consumer's and food information user's behavior. This allowed the identification of the main value components of FISs.FindingsThe authors propose a multidimensional framework for modelling the value proposition of digital platforms for FISs. Three main value dimensions have been identified: relevance, credibility, and accessibility. These dimensions concur with the consumer's perceived value of consumers in terms of benefits increase and cost reduction.Research limitations/implicationsThe research was intended to shed light on aspects characterizing consumers' perception of food information value. The authors put in evidence that the informational perspective of food communication is under-investigated. This study attempts to provide a holistic overview of the dimensions impacting on consumers' perception of the value of information for food consumers, opening new research perspectives.Practical implicationsThe framework represents a tool for positioning food information offerings on the market, with the objective to analyze the value proposition of FISs according to a consumer perspective and to understand gaps of current offering of FISs. Moreover, it may support the design of a new generation of digital platform for food information provision, which would respond to consumers' expectations and information needs, highlighting emerging business opportunities for digital entrepreneurs.Originality/valueFew research works provide a characterization of value proposition of digital platforms providing food information to consumers. In particular, to date, literature lacks of a holistic overview of the dimensions influencing consumer's perception of the information value of food communications.

UsbVisdaNet: User Behavior Visual Distillation and Attention Network for Multimodal Sentiment Classification.

In sentiment analysis, biased user reviews can have a detrimental impact on a company's evaluation. Therefore, identifying such users can be highly beneficial as their reviews are not based on reality but on their characteristics rooted in their psychology. Furthermore, biased users may be seen as instigators of other prejudiced information on social media. Thus, proposing a method to help detect polarized opinions in product reviews would offer significant advantages. This paper proposes a new method for sentiment classification of multimodal data, which is called UsbVisdaNet (User Behavior Visual Distillation and Attention Network). The method aims to identify biased user reviews by analyzing their psychological behaviors. It can identify both positive and negative users and improves sentiment classification results that may be skewed due to subjective biases in user opinions by leveraging user behavior information. Through ablation and comparison experiments, the effectiveness of UsbVisdaNet is demonstrated, achieving superior sentiment classification performance on the Yelp multimodal dataset. Our research pioneers the integration of user behavior features, text features, and image features at multiple hierarchical levels within this domain.

Research on Efficient Multi-Behavior Recommendation Method Fused with Graph Neural Network

Currently, most recommendation algorithms only use a single type of user behavior information to predict the target behavior. However, when browsing and selecting items, users generate other types of behavior information, which is important, but often not analyzed or modeled by traditional recommendation algorithms. This study aims to design a multi-behavior recommendation algorithm based on graph neural networks by analyzing multiple types of behavior information in users’ product purchasing process, to fully utilize multiple types of user behavior information. The algorithm models users, items, and user behavior in multiple dimensions by incorporating attention mechanisms and multi-behavior learning into graph neural networks, and solves the problem of imbalanced user behavior weights from the perspective of multi-task loss optimization. After experimental verification, we proposed that the multi-behavior graph attention network (MGAT) algorithm has better performance compared to four other classical recommendation algorithms on the Beibei and Taobao datasets. The results demonstrate that the multi-behavior recommendation algorithm based on graph neural networks has practicality in fully utilizing multiple types of user information, and can solve the problem of imbalanced user behavior weights to some extent.