Capture User Research Articles

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.

Where Do Users Look When Deciding If a Text Message is Safe or Malicious?

Phishing via SMS, or SMiShing, is a rapidly growing cyber threat that causes significant financial losses. While research on email phishing has explored user behavior, the understanding of human factors in SMiShing detection remains limited. This study bridges the gap by investigating how users visually evaluate real-world SMS message legitimacy using eye-tracking technology. We aim to identify which message components capture user attention (e.g., sender information, typos/grammatical errors, links) and assess if users focus sufficiently on established red flags associated with phishing attempts. This research contributes to more effective user-centric countermeasures against SMiShing attacks by informing the design of security interventions that consider user behavior and mobile device information limitations.

MSD: Multi-Order Semantic Denoising Model for Session-Based Recommendations

Session-based recommendations which aim to predict subsequent user–item interactions based on historical user behaviour during anonymous sessions can be challenging to carry out. Two main challenges need to be addressed and improved: (1) how does one analyze these sessions to accurately and completely capture users’ preferences, and (2) how does one identify and eliminate any interference caused by noisy behavior? Existing methods have not adequately addressed these issues since they either neglect the valuable insights that can be gained from analyzing consecutive groups of items or fail to take these noisy data in sessions seriously and handle them properly, which can jointly impede recommendation systems from capturing users’ real intentions. To address these two problems, we designed a multi-order semantic denoising (MSD) model for session-based recommendations. Specifically, we grouped items of different lengths into varying multi-order semantic units to mine the user’s primary intentions from multiple dimensions. Meanwhile, a novel denoising network was designed to alleviate the interference of noisy behavior and provide a more precise session representation. The results of extensive experiments on three real-world datasets demonstrated that the proposed MSD model exhibited improved performance compared with existing state-of-the-art methods in session-based recommendations.

Semantic distance of icons: Impact on user cognitive performance and a new model for semantic distance classification

With the continuous development of human-computer interaction technologies and the widespread use of graphical interfaces, icons that represent various objects and functions have a particularly important role. This study investigates the effect of semantic distance of icons on cognitive performance through an eye-movement-based experiment which involves a visual search for icons. The findings show that the semantic distance of icons has a significant effect on cognitive performance. A higher cognitive performance is found with semantically close icons which can better capture user attention. In addition, we use eye-movement indicators that are highly correlated with semantic distance, including mean pupil diameter, mean gaze duration and initial gaze time of an AOI, and analyze the objective relationship between these three eye-movement indicators and the semantic distance of icons to establish a dataset. The dataset is used as input for a Gradient Boosting Decision Tree (GBDT), which is a machine learning-based method for classifying the semantic distance of the icons in this study. The output of the GBDT model is classifying the semantic distance as far and close, and the experimental results show that the accuracy of the model reaches 84.28% after a comparison with other types of classifiers, which is in good agreement with the experimental results. Therefore, the model can address the relevant application requirements and simplify the evaluation process of icons to a certain extent, which has great significance in the field of icon design.

Identification of User Requirements and their Influencing Factors Based on Online Reviews and Operational Data

It is crucial for enterprises to clearly identify user needs during the process of formulating product design improvement plans. Therefore, it is essential to comprehensively and accurately identify user needs, explore the reasons behind the emergence of these needs, and incorporate user opinions into the process of product design improvement. A method is proposed to comprehensively and accurately capture user requirements and address the challenge of identifying the underlying causes of user requirements. This method utilizes online comments and operational data to identify user requirements and their influencing factors. First, text sentiment analysis techniques are employed to quantify the importance and performance values of product feature topics. Second, we construct a quadrant model to identify product features requiring improvement, and the original negative comments related to these features are traced. However, the quadrant model alone is insufficient to reflect specific product issues that users are concerned about. Therefore, a functional structure model based on product issues is designed to filter and identify factors that influence user requirements using operational data. Finally, a Bayesian network inference approach is utilized to identify the key influencing factors on user requirements, enabling analysis of the causes behind user requirements and the proposal of product design improvement strategies. The feasibility and effectiveness of the proposed method are validated through experiments conducted on heavy-duty truck data. By analyzing the original negative comments related to the power characteristics, specific user demands regarding the insufficient power of the product were identified, such as “obviously insufficient power when climbing slopes” and other issues. Based on the vehicle power system functional structure model, combined with expert knowledge and operational data, factors related to the state of parts and user behavior that may affect “insufficient vehicle power” were identified. Based on the analysis results, suggestions were made to improve the engine intake air temperature control strategy and to enhance vehicle performance by promoting correct user behavior through informational campaigns.

Towards Differential Privacy in Sequential Recommendation: A Noisy Graph Neural Network Approach

With increasing frequency of high-profile privacy breaches in various online platforms, users are becoming more concerned about their privacy. And recommender system is the core component of online platforms for providing personalized service, consequently, its privacy preservation has attracted great attention. As the gold standard of privacy protection, differential privacy has been widely adopted to preserve privacy in recommender systems. However, existing differentially private recommender systems only consider static and independent interactions, so they cannot apply to sequential recommendation where behaviors are dynamic and dependent. Meanwhile, little attention has been paid on the privacy risk of sensitive user features, most of them only protect user feedbacks. In this work, we propose a novel DIfferentially Private Sequential recommendation framework with a noisy Graph Neural Network approach (denoted as DIPSGNN) to address these limitations. To the best of our knowledge, we are the first to achieve differential privacy in sequential recommendation with dependent interactions. Specifically, in DIPSGNN, we first leverage piecewise mechanism to protect sensitive user features. Then, we innovatively add calibrated noise into aggregation step of graph neural network based on aggregation perturbation mechanism. And, this noisy graph neural network can protect sequentially dependent interactions and capture user preferences simultaneously. Extensive experiments demonstrate the superiority of our method over state-of-the-art differentially private recommender systems in terms of better balance between privacy and accuracy.

Alleviating Video-length Effect for Micro-video Recommendation

Micro-video platforms such as TikTok are extremely popular nowadays. One important feature is that users no longer select interested videos from a set; instead, they either watch the recommended video or skip to the next one. As a result, the time length of users’ watching behavior becomes the most important signal for identifying preferences. However, our empirical data analysis has shown a video-length effect that long videos can more easily receive a higher value of average view time, and thus adopting such view-time labels for measuring user preferences can easily induce a biased model that favors the longer videos. In this article, we propose a V ideo L ength D ebiasing Rec ommendation (VLDRec) method to alleviate such an effect for micro-video recommendation. VLDRec designs the data labeling approach and the sample generation module that better capture user preferences in a view-time-oriented manner. It further leverages the multi-task learning technique to jointly optimize the above samples with the original biased ones. Extensive experiments show that VLDRec can improve users’ view time by 1.81% and 11.32% on two real-world datasets, given a recommendation list of a fixed overall video length, compared with the best baseline method. Moreover, VLDRec is also more effective in matching users’ interests in terms of the video content.

Multi-aspect features of items for time-ordered sequential recommendation

The key to sequential recommendation modeling is to capture dynamic users’ interests. Existing sequential recommendation methods (e.g., self-attention mechanism) have achieved extraordinary success in modeling users’ interests. However, these models ignore that users have different levels of preferences for different aspects of items, failing to capture users’ most concerning aspects. In addition, they are highly dependent on the quality of training data, which may lead to overfitting of the model when the training data is insufficient. To address the above issues, we propose a novel sequence-aware model (Multi-Aspect Features of Items for Time-Ordered Sequential Recommendation, MFITSRec), which combines the features of items with user behavior sequences to learn more complex item-item and item-attribute relationships. Moreover, the model uses a self-attention network based on an absolute time relationship, which can better represent the changes in users’ interests and capture users’ preferences for particular aspects of items. Extensive experiments on five datasets demonstrate that our model outperforms various baseline models. In particular, the model’s prediction accuracy has been significantly improved on sparse datasets.

Heterogeneous Hypergraph Neural Network for Social Recommendation using Attention Network

Graph neural networks (GNNs) have been used extensively as a backbone for social recommendation. However, their application to a diverse range of situations is still rather limited. This is because graph structures only leverage pairwise user relationships. They cannot capture the higher-order relationships so common in the real world, and ignoring the interest friends and strangers might have in similar items is severely hampering the expressiveness of the current graph-based recommendation models. Hence, in this paper, we outline a heterogeneous hypergraph neural network for social recommendation called HHGSA that incorporates an attention network to address these issues. The hypergraph is able to represent higher-order relationships through five motifs: friend and stranger item appeal, item similarity, user similarity based on interactions with items, and social relations. Two modules, the attentive vertex aggregation module, and the attentive hyperedge aggregation module, capture user and item attention. In addition, it has been discovered that similar items have identical appeal when displayed to users. A GNN aggregates the user embedding data, including information about the friend and stranger and item embeddings. Finally, information about users and items is aggregated for social recommendations. Extensive experiments on four datasets demonstrate that the HHGSA model outperforms a wide range of baselines and can significantly improve the accuracy of recommendations.

SMAR: Summary-Aware Multi-Aspect Recommendation

Extracting user preferences and item features from reviews to assist recommendations is becoming increasingly popular. However, on the one hand, existing works generally select reviews based on how well user reviews match item reviews. They ignore that reviews may contain noise such as irrelevant phrases, which will affect the accuracy of selecting important reviews. In contrast, summaries written by users are abstracts of reviews that contain critical item feature information. They can be adopted to identify crucial reviews and further capture user’s fine-grained preferences from reviews. In addition, current methods do not consider that different items have different aspects in the same domain. They normally set a fixed number of aspects of the entire domain to get coarse-grained user preferences and item features. However, when modeling the user’s preferences for the current item, it might be more important to capture the corresponding aspects of the item preferences. Therefore, in this paper, we are motivated to propose a Summary-Aware Multi-Aspect Recommendation (SMAR). Specifically, we first construct a Summary-Aware Review Selection Module which adopts summaries to alleviate noise in reviews, identifying key reviews accurately. We then design a Summary-Aware Multi-Aspect Module which captures targeted user preferences towards the current item’s aspects. Finally, we employ Latent Factor Model to complete the recommendation process. The experimental results on Amazon datasets show that our method significantly outperfoms state-of-art approaches in terms of rating prediction accuracy.

Self-Supervised Interest Transfer Network via Prototypical Contrastive Learning for Recommendation

Cross-domain recommendation has attracted increasing attention from industry and academia recently. However, most existing methods do not exploit the interest invariance between domains, which would yield sub-optimal solutions. In this paper, we propose a cross-domain recommendation method: Self-supervised Interest Transfer Network (SITN), which can effectively transfer invariant knowledge between domains via prototypical contrastive learning. Specifically, we perform two levels of cross-domain contrastive learning: 1) instance-to-instance contrastive learning, 2) instance-to-cluster contrastive learning. Not only that, we also take into account users' multi-granularity and multi-view interests. With this paradigm, SITN can explicitly learn the invariant knowledge of interest clusters between domains and accurately capture users' intents and preferences. We conducted extensive experiments on a public dataset and a large-scale industrial dataset collected from one of the world's leading e-commerce corporations. The experimental results indicate that SITN achieves significant improvements over state-of-the-art recommendation methods. Additionally, SITN has been deployed on a micro-video recommendation platform, and the online A/B testing results further demonstrate its practical value. Supplement is available at: https://github.com/fanqieCoffee/SITN-Supplement.

Understanding how indoor environmental classroom conditions influence academic performance in higher education

PurposeThis study aims to qualitatively examine the relationship between the indoor environmental quality (IEQ), lecturers’ and students’ perceived internal responses and academic performance.Design/methodology/approachTo capture user experiences with the IEQ in classrooms, semi-structured interviews with 11 lecturers and three focus group discussions with 24 students were conducted, transcribed, coded and analyzed using direct content analysis.FindingsThe findings show that lecturers and students experience poor thermal, lighting, acoustic and indoor air quality (IAQ) conditions that may influence their ability to teach and learn. Maintaining acceptable thermal and IAQ conditions was difficult for lecturers, as opening windows or doors caused noise disturbances. In uncomfortable conditions, lecturers may decide to give a break earlier or shorten a lecture. When students experienced discomfort, it may affect their ability to concentrate, their emotional status and their quality of learning.Research limitations/implicationsThe findings originate from a relatively small sample, which might have limited the number and variety of identified associations between environment and users.Practical implicationsMaintaining acceptable air and thermal conditions will mitigate the need to open windows and doors. Keeping doors and windows closed will prevent noise disturbances and related distractions. This will support the quality of learning in classrooms. This study reveals the end users’ perspectives and preferences, which can inspire designers of new school buildings in higher education.Originality/valueThis study emphasizes the importance of creating and maintaining optimal IEQ conditions to support the quality of teaching and learning. These conditions are particularly relevant when classroom occupancy rates are high or outdoor conditions are unfavourable.

Complaint Box App Implementation Using React Native

Abstract: This paper presents the implementation of a complaint box app using React Native, a cross-platform mobile app development framework. The app allows users to submit complaints and suggestions, while the admin panel enables complaint handlers to manage complaints and communicate with users. The app features user authentication, complaint submission, complaint tracking, and real-time communication, while the implementation involved designing the user interface, integrating it with a backend system and a database, and testing its functionality. The paper discusses the benefits and challenges of using React Native for mobile app development and how it can be leveraged to build scalable and efficient complaint box apps. Overall, this paper presents a practical and effective approach to implementing a complaint box app using React Native with an admin panel, which can help organizations capture user feedback and improve their products and services.

Knowledge-Aware Enhanced Network Combining Neighborhood Information for Recommendations

In recent years, the application of knowledge graphs to alleviate cold start and data sparsity problems of users and items in recommendation systems, has aroused great interest. In this paper, in order to address the insufficient representation of user and item embeddings in existing knowledge graph-based recommendation methods, a knowledge-aware enhanced network, combining neighborhood information recommendation (KCNR), is proposed. Specifically, KCNR first encodes prior information about the user–item interaction, and obtains the user’s different knowledge neighbors by propagating them in the knowledge graph, and uses a knowledge-aware attention network to distinguish and aggregate the contributions of the different neighbors in the knowledge graph, as a way to enrich the user’s description. Similarly, KCNR samples multiple-hop neighbors of item entities in the knowledge graph, and has a bias to aggregate the neighborhood information, to enhance the item embedding representation. With the above processing, KCNR can automatically discover structural and associative semantic information in the knowledge graph, and capture users’ latent distant personalized preferences, by propagating them across the knowledge graph. In addition, considering the relevance of items to entities in the knowledge graph, KCNR has designed an information complementarity module, which automatically shares potential interaction characteristics of items and entities, and enables items and entities to complement the available information. We have verified that KCNR has excellent recommendation performance through extensive experiments in three real-life scenes: movies, books, and music.

A Comparative Analysis of Sampling Techniques for Click-Through Rate Prediction in Native Advertising

Native advertising is a popular form of online advertisements that has similar styles and functions with the native content displayed on online platforms, such as news, sports and social websites. It can better capture users’ attention, and they have gained increasing popularity in many online platforms and among advertisers. In advertising, Click Trough Rate (CTR) prediction is essential but challenging due to data sparsity: the non-clicks constitute most of the data, whereas clicks form a significantly smaller portion. The performance of 19 class imbalance approaches is compared in this study with the use of four traditional classifiers, to determine the most effective imbalance methods for our native ads dataset. The data used is real traffic data from Finland over the course of seven days provided by the native advertising platform ReadPeak. The resampling methods used include seven undersampling techniques, four oversampling techniques, four hybrid sampling techniques, and four ensemble systems. The findings demonstrate that class imbalance learning can enhance the model’s capacity for classification by as much as 20%. In general, oversampling is more stable comparatively. But, undersampling performed the best with Random Forest. Our study also demonstrates that the imbalance ratio plays an important role in the performance of the model and the features importance.

HyperRS: Hypernetwork-Based Recommender System for the User Cold-Start Problem

Meta-learning has been proven to be effective for the cold-start problem of recommender systems. Many meta-learning recommender systems that are designed for the user cold-start problem are gradient-based. They use a global parameter learned from existing users to initialize the recommender system parameter for new users that provides a personalized recommendation with limited user-item interactions. Many systems require users’ demographic information to learn the global parameter. This requirement raises privacy concerns, and demographic information is not always available. In addition, some of the gradient-based systems need dozens or even hundreds of user-item interactions from existing users to learn the global parameter. This requirement is difficult to satisfy in specific scenarios in which active users and their item interactions are rare. Moreover, gradient-based meta-learning systems rarely capture user preferences over different item attributes, as updating the corresponding weights requires many optimization steps. We proposed HyperRS, a Hypernetwork-based Recommender System, for the user cold-start problem. Our system does not rely on demographic information to provide personalized recommendations. In our system, a hypernetwork generates all weights in the underlying recommender system. The hypernetwork enables the weights to adapt quickly to capture user interest in both item attributes and the contents of the item attributes. The experimental results show that our method outperforms several state-of-the-art meta-learning recommender systems for the user cold-start problem.

Distributed fuzzy clustering algorithm for mixed-mode data in Apache SPARK

Fuzzy clustering is an invaluable data mining technique that allows each data point to belong to more than one cluster with some degree of membership. It is widely employed in exploratory data mining to discover overlapping communities in social networks, find structure in spectral data, and capture user interests in recommendation systems. Nowadays, the variety and volume of data are increasing at a tremendous rate. Data is power; the massive data, along with an effective technique, can unravel valuable information. The existing fuzzy clustering algorithms do not perform well on massive heterogeneous datasets. Processing an enormous amount of data is beyond the capacity of a single processor. The need of the hour is to develop fuzzy clustering techniques that can work on a distributed framework for Big Data processing and can handle heterogeneous data. In this research, we evaluate the performance of the recently proposed algorithm for the Fuzzy clustering of mixed-mode data FCMD-MD (D’Urso and Massari in Inf Sci 505:513–534, 2019) with different real-world datasets. We develop a distributed FCMD-MD, a fuzzy clustering algorithm for mixed-mode data in Apache SPARK. The experimental results show that the algorithm is scalable, performs well in a distributed environment, and clusters enormous heterogeneous data with high accuracy. We also compared the performance of distributed FCMD-MD and the distributed k-medoid algorithm.

Dynamic clustering based contextual combinatorial multi-armed bandit for online recommendation

Recommender systems still face a trade-off between exploring new items to maximize user satisfaction and exploiting those already interacted with to match user interests. This problem is widely recognized as the exploration/exploitation (EE) dilemma, and the multi-armed bandit (MAB) algorithm has proven to be an effective solution. As the scale of users and items in real-world application scenarios increases, their purchase interactions become sparser. Then three issues need to be investigated when building MAB-based recommender systems. First, large-scale users and sparse interactions increase the difficulty of user preference mining. Second, traditional bandits model items as arms and cannot deal with ever-growing items effectively. Third, widely used Bernoulli-based reward mechanisms only feedback 0 or 1, ignoring rich implicit feedback such as behaviors like click and add-to-cart. To address these problems, we propose an algorithm named Dynamic Clustering based Contextual Combinatorial Multi-Armed Bandits (DC3MAB), which consists of three configurable key components. Specifically, a dynamic user clustering strategy enables different users in the same cluster to cooperate in estimating the expected rewards of arms. A dynamic item partitioning approach based on collaborative filtering significantly reduces the scale of arms and produces a recommendation list instead of one item to provide diversity. In addition, a multi-class reward mechanism based on fine-grained implicit feedback helps better capture user preferences. Extensive empirical experiments on three real-world datasets demonstrate the superiority of our proposed DC3MAB over state-of-the-art bandits (On average, +75.8% in F1 and +54.3% in cumulative reward). The source code is available at https://github.com/HaixHan/DC3MAB.

Ask the Users

Just-in-Time Adaptive Interventions (JITAIs) are envisioned to harness rich data on users' contexts. However, many JITAIs fall short in leveraging the value of the data while sculpting the interventions. Investigating the literature reveals a lack of user-centered design (UCD) methods in designing JITAIs. Our case study of applying a UCD process revealed that even without deploying a JITAI, UCD could uncover user interactions that inform critical design decisions of a JITAI's components. We reflect on our experiences engaging in a user-centered JITAI design process and urge the broader human-computer interaction (HCI) community to devise concrete design guidelines for JITAIs. We take the first step toward that goal by proposing a checklist of key design considerations for future JITAIs. Together, this case study contributes insights on applying UCD to the design of all components of a JITAI, which can help capture users' needs and provide actionable interventions.

Full-Body Motion Capture-Based Virtual Reality Multi-Remote Collaboration System

Various realistic collaboration technologies have emerged in the context of the COVID-19 pandemic. However, as existing virtual reality (VR) collaboration systems generally employ an inverse kinematic method using a head-mounted display and controller, the user and character cannot be accurately matched. Accordingly, the immersion level of the VR experience is low. In this study, we propose a VR remote collaboration system that uses motion capture to improve immersion. The system uses a VR character in which a user wearing motion capture equipment performs the same operations as the user. Nevertheless, an error can occur in the virtual environment when the sizes of the actual motion capture user and virtual character are different. To reduce this error, a technique for synchronizing the size of the character according to the user’s body was implemented and tested. The experimental results show that the error between the heights of the test subject and virtual character was 0.465 cm on average. To verify that the implementation of the motion-capture-based VR remote collaboration system is possible, we confirm that three motion-capture users can collaborate remotely using a photon server.