User Features Research Articles

Enhancing Biometric Security with Bimodal Deep Learning and Feature-level Fusion of Facial and Voice Data

Recent research in biometric technologies underscores the benefits of multimodal systems that use multiple traits to enhance security by complicating the replication of samples from genuine users. To address this, we present a bimodal deep learning network (BDLN or BNet) that integrates facial and voice modalities. Voice features are extracted using the SincNet architecture, and facial image features are obtained from convolutional layers. Proposed network fuses these feature vectors using either averaging or concatenation methods. A dense connected layer then processes the combined vector to produce a dual-modal vector that encapsulates distinctive user features. This dual-modal vector, processed through a softmax activation function and another dense connected layer, is used for identification. The presented system achieved an identification accuracy of 99% and a low equal error rate (EER) of 0.13% for verification. These results, derived from the VidTimit and BIOMEX-DB datasets, highlight the effectiveness of the proposed bimodal approach in improving biometric security.

UeLP: accurate user linkage across social platforms against location errors

Abstract User linkage across social platforms can connect the accounts of the same user across different social networks, which is crucial for the identification of users’ multiple social identities and cross-platform association analysis. Cross-platform user linkage based on location is a typical method in current research. These methods typically rely on check-in data to calculate user similarity. However, different from check-in location, the location data obtained from instant messaging social platforms may contain random errors, leading to low accuracy of user linkage of such methods. To solve this problem, this paper proposes an accurate user linkage method across social platforms against location errors. First, unlike existing methods that employ fixed-size grids, this paper uses a multi-grained spatio-temporal grid to organize data, in order to accurately extract user features from error locations. Then, by extracting coarse-grained movement pattern features from user trajectories, candidate users are filtered out, and a small subset of candidate uses is generated to effectively reduce the search space. Next, we establish a weight model based on grid contribution and motion sequence similarity to extract location and temporal features with stronger user orientation. Finally, according to the weight model, the weighted cluster center distance of trajectories is used to calculate the similarity between two different user trajectories. The user with the highest similarity is selected from the candidate subset to complete the user linkage. The extensive experiments are conducted on six public datasets containing 115 866 trajectories and a self-built dataset with 5358 trajectories. The results show the following: compared with the four existing typical location-based methods $k$-BCT, GS, TF-IDF, and TF-IWF, the accuracy Acc@1 is improved by an average of 33%, 44.94%, 15.2%, and 14.55%, respectively, and the accuracy Acc@3 is improved by 30.52%, 34.67%, 13.84%, and 13.19%, respectively.

Combining style generative adversarial networks with particle swarm optimisation-support vector regression to design affective social robot for public health intervention

In order to improve the attractiveness of social robot serving health interventions in public workspace, we propose a product design method with the combination of Style-generative adversarial network (StyleGAN) model and particle swarm optimisation-support vector regression (PSO-SVR). This paper aims to explore the modelling generation method of social robots for health intervention based on artificial intelligence generated content (AIGC) and a mapping model between the shape characteristics and users’ visual perception based on Kansei Engineering (KE). Firstly, to address the defects of typical KE over-reliance on existing samples, we introduce the StyleGAN model in AIGC to learn and train existing robots’ shape samples and generate new robots’ shape sample images. Secondly, the morphological deconstruction method is used to deconstruct the shape features of the new robot sample. Factor analysis (FA) is also used to reduce dimension and cluster emotional words and establish a Likert scale between the shape features of robots and users’ emotional vocabulary. Finally, particle swarm optimisation-support vector regression (PSO-SVR) is used to establish a mapping model between the shape features of social robots and the emotional images of users, thus obtaining the most attractive shape combination scheme. The research results showed that the StyleGAN model in AIGC can be used to assist industrial designers in creative expression and provide rich sample sources for KE; and the PSO-SVR of machine learning can be used to build a mapping model among users’ visual feelings, emotional images and shape features. In the end, we designed an attractive social robot for public health intervention.

Analysis of Software Developers' Programming Language Preferences and Community Behavior From Big5 Personality Traits

ABSTRACTMany programming languages and technologies have appeared for the purpose of software development. When choosing a programming language, the developers' cognitive attributes, such as the Big5 personality traits (BPT), may play a role. The developers' personality traits can be reflected in their social media content (e.g., tweets, statuses, Q&A, reputation). In this article, we predict the developers' programming language preferences (i.e., the pattern of picking up a language) from their BPT derived from their content produced on social media. We randomly collected data from a total of 820 Twitter (currently X) and Stack Overflow (SO) users. Then, we collected user features (i.e., BPT, word embedding of tweets) from Twitter and programming preferences (i.e., programming tags, reputation, question, answer) from SO. We applied various machine learning (ML) and deep learning (DL) techniques to predict their programming language preferences from their BPT. We also investigated other interesting insights, such as how reputation and question‐asking/replying are associated with the users' BPT. The findings suggest that developers with high openness, conscientiousness, and extraversion are inclined to mobile applications, object‐oriented programming, and web programming, respectively. Furthermore, developers with high openness and conscientiousness traits have a high reputation in the SO community. Our ML and DL techniques classify the developers' programming language preferences using their BPT with an average accuracy of 78%.

CBRec: A causal way balancing multidimensional attraction effect in POI recommendations

In the next Point-of-Interest recommendation, sparse and uneven location data generate biases, resulting in homogeneous recommendation outcomes that fail to reflect user preferences. Although there are many related unbiased studies, they still exhibit limitations. They lack a unified debiasing paradigm and typically employ different methods to address various biases, resulting in complex and incompatible debiasing models. Additionally, they often overlook the potential advantages of biases, thus harming the quality of location features. To address these challenges, we propose a unified debiasing paradigm by intervening in location attraction to balance the positive and negative effects of bias. By analyzing the structural causal graph, we identify attraction as a feature influenced by bias. By comparing observational results affected by attraction with counterfactual results unaffected by it, we derive a unified debiasing paradigm that eliminates the effects of bias. Additionally, through feature fusion, we embed multidimensional attraction into user features, leveraging the advantages of bias to preserve the quality of location features. Finally, experimental results on five real-world datasets demonstrate that our proposed model outperforms recent sequential recommendation models.

Social robot detection based on user behavioral representation

Social robots aim to facilitate our lives by mimicking our behaviors to deliver information and interact with us. Unfortunately, some social robots are bred for sinister intention, which often spread fake news, even threatening personal and national security. To effectively detect these robots, some existing works utilize basic user attributes for clustering, such as creation time and other information, but ignore the behavioral features of social robots, which are important a priori information for detection, and their absence leads to poor performance. To address this problem, we construct the Social Robot Behavior Detection model, which employs social robots' behavioral habits in disseminating content. Specifically, our method consists of three parts. First, to accurately characterize news, we analyze news features from multiple views, which facilitates representing users' habits of posting news. Second, a novel aggregation strategy is designed for aggregating user features to obtain accurate user representation. Finally, a classifier is used to detect whether user is social robots or not. We conduct analytical experiments on two public datasets, and the results show that our model has better detection performance, which demonstrates that our approach outperforms existing works on social robot detection. Our source code is available at https://github.com/NPURG/SRDB.

Research on modeling method for digitizing distribution substation areas based on big data from smart meters

With the wide application of smart meters, real-time data collection in power grid operation can be carried out in real time, which provides big data support for the digital modeling of distribution station areas. At the same time, the digital transformation of the distribution area can be promoted through the digital modeling of the distribution area. In this situation, how to use the big data of smart meters to achieve the digital modeling of the distribution area needs to be further studied. Firstly, this paper briefly introduces the connotation of digitalization in the distribution station area. At the same time, aiming at the digitalization of the distribution station area, it analyzes the digitalization of station features, user features, new energy features, network parameters and operation features, and obtains the digitalization model. Finally, on the basis of the above, select an application scenario to introduce the application of digitalization modeling. The research results can provide a reference for the combination of big data application of smart meters and digitalization of the distribution area, and help the digital transformation of the distribution area.

A two-stage recommendation optimization algorithm based on item popularity and user features

Financial product recommendation algorithms are mainly product-centered. This article proposes a two-stage recommendation optimization algorithm based on item popularity and user features, named CPCF-TSP, that can make full use of the demographic characteristics of users and mitigate the problem of users being more inclined to choose “hot” financial products. A popularity weight factor is introduced to normalize popularity and modify Pearson's similarity function. The modified Pearson's similarity function is combined with popularity normalization and user features to improve modeling performance. The two-stage recommendation optimization procedure was combined with a collaborative filtering algorithm to improve recommendation precision. CPCF-TSP fully considers user features in building a hybrid recommendation model and solves the problem of user cold-start. It can also mitigate popularity deviations and improve recommendation precision. MovieLens data and Santander Bank client trading data were used in a case study. The results show that the algorithm reduces inaccuracy in the calculation of the weights for recommendation popularity and similarity and is especially suitable for recommending financial products in which user information can be easily collected and the number of users is far greater than the number of products considered.

Reshaping of Global Shipping Dynamics: Implications for Strategic Resilience Capabilities, Marketing Capabilities, Supply Chain Orientation Capabilities and Port Competitiveness in Indonesia

The phenomenon of dynamic maritime environment have affected port business and it is substantially changing the business environment in which large-scale container trade growth is being facilitated. In this paper will illustrate the dynamic features of port users, discuss the port competition strategy and analyse the marketing strategic for port business under transformation phases. Using qualitative methods using case studies through a literature review and empirical evidence to prove, deepen and expand the results of the study. In this article, we develop and examine the implication such marketing capabilities approach to the effectiveness of the marketing strategies and strategic resilience capabilities to determine sustainable competitive performance. Based on a discussion among with its managements, this paper could prove a conceptual framework in which strategic resilience capabilities and supply chain orientation capabilities have a profound impact on the company’s sustainable competitive performance.

Publisher preferences for a journal transparency tool: A modified three-round Delphi study

Background We propose the creation of a journal transparency tool (JTT), which will allow users to obtain information about a given scholarly journal’s operations and policies. We are obtaining preferences from different stakeholders to inform the development of this tool. This study aimed to identify the publishing community’s preferences for the JTT. Methods We conducted a modified three-round Delphi survey. Representatives from publishing houses and journal publishers were recruited through purposeful and snowball sampling. The first two Delphi rounds involved an online survey with items about JTT metrics and user features. During the third round, participants discussed and voted on JTT metric items that did not reach consensus after round 2 within a virtual consensus meeting. We defined consensus as 80% agreement to include or exclude an item in the JTT. Results Eighty-six participants completed the round 1 survey, and 43 participants (50% of round 1) completed the round 2 survey. In both rounds, respondents voted on JTT user feature and JTT metric item preferences and answered open-ended survey questions regarding the JTT. In round 3, a total of 21 participants discussed and voted on JTT metric items that did not reach consensus after round 2 during an online consensus group meeting. Fifteen out of 30 JTT metric items and none of the four JTT user feature items reached the 80% consensus threshold after all rounds of voting. Analysis of the round 3 online consensus group transcript resulted in two themes: ‘factors impacting support for JTT metrics’ and ‘suggestions for user clarity.’ Conclusions Participants suggested that the publishing community’s primary concerns for a JTT are to ensure that the tool is relevant, user-friendly, accessible, and equitable. The outcomes of this research will contribute to developing and refining the tool in accordance with publishing preferences.

Readability and Accessibility of Patient-Education Materials for Heart Failure in the United States

Online education materials are widely used by patients and caregivers to understand the management of complex chronic diseases such as heart failure (HF). Organizations such as the American Medical Association and National Institutes of Health recommend that materials be written at a 6th grade reading level. The current study examined the readability and accessibility of online education materials for patients with HF. Whole page text from each included website was entered into an online readability calculator. Five validated readability indices (Flesch-Kincaid Grade Level, Flesch Reading Ease Scale, Gunning Fog Index, Coleman-Liau Index, and Simple Measure of Gobbledygook (SMOG Index)) were used to evaluate each source. Websites were categorized by source (government, public, and private). The availability of audiovisual accessibility features and content in non-English languages were assessed for each website. Of the 36 online resources analyzed, the median readability level was 9-10th grade by the Flesch-Kincaid Grade Level and college level using the Flesch Reading Ease Scale. The Gunning Fog Index and Coleman-Liau Index both showed median readability scores corresponding to a 12th grade reading level, while the SMOG Index showed a median score corresponding to that of the 9th grade. Only 10 websites (28%) offered information in languages other than English, and none provided comprehensive accessibility features for users with disabilities. Common online educational materials for patients with HF are characterized by a higher readability level than that recommended by the National Institutes of Health and American Medical Association with limited multilingual and accessibility options, potentially limiting the accessibility of resources to patients and caregivers.

Assessing the visual landscape of Istanbul Bosphorus: Exploring the role of vegetation and built environment characteristics

Istanbul Bosphorus is one of the most important protected landscapes of Türkiye with its dramatic topography, natural landscape, and historical and cultural heritage embedded within this natural setting. However, studies focusing on the visual landscape assessment (VLA) of Bosphorus are limited. The main goal was to evaluate the VLA of Bosphorus and to understand the users' perception of this multi-layered landscape via utilizing VLA based on the evaluation of 45 photographs taken from the sea. For this goal, the study investigates the environmental texture proportion (ETP: vegetation, built environment, sea, and sky) and RGB values of the photographs using parameters such as naturalness, beautiful scenery, disturbance, excitement, safety, and livability in 5-Likert’s scale. Findings demonstrated that photographs with more vegetation than built environment, with those containing 10 % or more vegetation receiving higher overall scores. Vegetation showed a high positive correlation with naturalness, beautiful scenery, excitement, safety, and livability. Moreover, disturbance was positively correlated with the built environment, and negatively correlated with vegetation. Namely, the most problematic ETP type for VLA was the built environment in Bosphorus. The study further identified built environment characteristics as a critical environmental component, with historical and mansion-type buildings scoring higher than other building types. Another important finding was that certain user features, such as gender, influenced perceptions of beautiful scenery (9 p-value), excitement (10), safety (11), and livability (21), while usage frequency affected perceptions of excitement (14) and safety (13). In other words, men gave higher scores to photographs compared to women, and users who regularly utilized the study area rated the photographs more favorably than those who used it less frequently. In conclusion, the study found that user characteristics of gender and usage frequency were influential factors, and vegetation (10 % and above) was the most paramount element for preferences when considering the visual landscape of the Bosphorus in Istanbul.

TB-BGAT With TinyBERT and BiGRU in Personalized Course Recommendations

Aiming at the problems of inadequate user feature extraction, cold start and sparse data, a personalized course recommendation algorithm that utilizes TB-BGAT is suggested. First, the Tiny Bidirectional Encoder Representation from Transformers (TinyBERT) model is utilized to output character-level word vectors; then, Bidirectional Recurrent Neural Network (BiGRU) model is utilized to obtain the embedded contextual semantic features. Finally, the attention mechanism is utilized to allocate weights to various course features by assigning their importance and to obtain the output results. The results of experiment on the publicly available dataset MOOCs-Course prove that the proposed method improves at least 3.62%, 3.04%, and 3.33% in precision, recall, and F1-score, correspondingly, in contrast to several other state-of-the-art course resource recommendation algorithms. The proposed method can enhance the effectiveness of the course recommendation model, enhance the quality of learners' online learning, and provide good technical support for online education learning platforms.

Click-through rate prediction model based on graph networks and feature squeeze-and-excitation mechanism

PurposePredicting a user’s click-through rate on an advertisement or item often uses deep learning methods to mine hidden information in data features, which can provide users with more accurate personalized recommendations. However, existing works usually ignore the problem that the drift of user interests may lead to the generation of new features when they compute feature interactions. Based on this, this paper aims to design a model to address this issue.Design/methodology/approachFirst, the authors use graph neural networks to model users’ interest relationships, using the existing user features as the node features of the graph neural networks. Second, through the squeeze-and-excitation network mechanism, the user features and item features are subjected to squeeze operation and excitation operation, respectively, and the importance of the features is adaptively adjusted by learning the channel weights of the features. Finally, the feature space is divided into multiple subspaces to allocate features to different models, which can improve the performance of the model.FindingsThe authors conduct experiments on two real-world data sets, and the results show that the model can effectively improve the prediction accuracy of advertisement or item click events.Originality/valueIn the study, the authors propose graph network and feature squeeze-and-excitation model for click-through rate prediction, which is used to dynamically learn the importance of features. The results indicate the effectiveness of the model.

Multi-stream graph attention network for recommendation with knowledge graph

In recent years, the powerful modeling ability of Graph Neural Networks (GNNs) has led to their widespread use in knowledge-aware recommender systems. However, existing GNN-based methods for information propagation among entities in knowledge graphs (KGs) may not efficiently filter out less informative entities. To address this challenge and improve the encoding of high-order structure information among many entities, we propose an end-to-end neural network-based method called Multi-stream Graph Attention Network (MSGAT). MSGAT explicitly discriminates the importance of entities from four critical perspectives and recursively propagates neighbor embeddings to refine the target node. Specifically, we use an attention mechanism from the user's perspective to distill the domain nodes' information of the predicted item in the KG, enhance the user's information on items, and generate the feature representation of the predicted item. We also propose a multi-stream attention mechanism to aggregate user history click item's neighborhood entity information in the KG and generate the user's feature representation. We conduct extensive experiments on three real datasets for movies, music, and books, and the empirical results demonstrate that MSGAT outperforms current state-of-the-art baselines.

Functional optimization of the design of adaptive furniture to improve the accessibility and comfort of people with impaired musculosical system

In today's world, there is a growing need to create an inclusive environment that meets the needs of people with limited mobility. This applies not only to architecture and public spaces at the level of macro- and meso-ergonomics, but also to adaptive furniture at the level of micro-ergonomics. The study is devoted to the formalization of a number of requirements for the modeling of adaptive furniture for people with musculoskeletal disorders. Particular attention is paid to aspects in the development of furniture that not only provide comfort, safety and functionality, but also represent an aesthetically pleasing design that contributes to the integration of ASPA into the social environment. Ergonomic requirements are considered, taking into account the anthropometric and biomechanical features of such users with the help of digital models (DHM). Using DHM, you can study the interaction of users with furniture, simulate different scenarios of their use, and evaluate the parameters of the quality of comfort and safety. It has been established that the integration of DHM with modern three-dimensional modeling technologies allows optimizing furniture designs, making them maximally adapted to the needs of ASPA. The purpose of the study is to develop a methodology for designing adaptive furniture, which ensures an increase in the standard of living of users thanks to individually configured ergonomic. In order for ASPAs to be able to work comfortably and productively, the projected development of the interpenetration of modern technical innovations in this field with the subsequent unification in a separate direction of designing inclusive furniture and equipment, with the division into separate groups for different types of disabilities, becomes quite relevant.

KGCF: Social relationship-aware graph collaborative filtering for recommendation

With the increasing popularity of recommendation techniques and social networks, social network recommendation has become a significant research field, i.e., predicting a user's preferences based on her or his historical interaction data, because the social relationships of users can not only enrich their interaction information, but also imply the accurate characterization of their preferences. Although there are some existing studies that predict users' preferences by using the characteristics of social relationships, their time complexity and hardware resource consumption are often very expensive, which limits the feasibility in large-scale real-world scenarios. To address these issues, in this paper, we propose a simple and effective K-core Graph Collaborative Filtering (KGCF) model to incorporate the user's social features into a recommendation framework. The user's interaction information is auto-encoded linearly by incorporating the social relationship graph into the user's interaction information and constructing a multi-layer graph filter. The linear autoencoder graph filter alleviates the highly sparse data problem and dramatically reduces the training time and hardware resource consumption. Experimental results on several real-world datasets show the superior effectiveness and spatiotemporal efficiency compared with the existing baselines, especially the training speed is significantly accelerated for large-scale datasets.

SYSTEM DESIGN OF INTERACTIVE PRODUCTS IN URBANISM

The article analyses the issue of a systematic approach to the design of digital products that form a modern urban space, forming a “media city”. It is shown that in modern cities digital space plays a key role in the development of the concept of “smart cities”, and digital technologies are increasingly used to solve urban problems and improve the quality of life of citizens. The characteristic features of user interfaces, such as cross-platform and ecosystem, are highlighted and described, as well as the topic of cultural, aesthetic and semantic aspects of design is touched upon. The practical experience of educational design in the field of interactive solutions for the urban environment at the Department of Environmental Design of Stroganov Russian State University of Design and Applied Arts is summarised. Instructional design experiences have been shown to add an experimental approach to systemic problem solving.

Design of a news recommendation model based on collaborative filtering and LSTM

This paper aims to design a news recommendation model based on user preferences to address the issues in recommendation systems under large datasets. Initially, four datasetsclick_history, news, news_embedding, and user_predictwere integrated into a single table, followed by data cleaning and feature engineering. Due to the large volume of data, this paper proposes necessary data filtering for the training and testing sets, utilizing temporal data to construct user feature vectors and news feature vectors. One challenge is how to effectively integrate user preferences and news features into the model to avoid overfitting or underfitting. In the model design and building phase, different methods were attempted to merge the information of users and news. Ultimately, the user preference features were processed using a fully connected layer, and the news embedding vectors were handled using an LSTM model. These two data parts were then combined into another fully connected layer, using ReLU as the activation function and CELoss as the loss function. Subsequently, the model's hyperparameters were adjusted and evaluated, achieving favorable model performance. The prediction accuracy for recommending news to users in user_predict was calculated as an evaluation criterion. Finally, this paper proposes directions for generalization and optimization in three aspects: data processing, model design, and experimental design. This includes data processing methods, potential improvements or mechanisms that could be incorporated into the model, and hyperparameter tuning. The paper primarily proposes data filtering to solve the problem of excessive data scale, which may aid in addressing recommendation system issues under large-scale datasets.

Deep Causal Reasoning for Recommendations

Traditional recommender systems aim to estimate a user’s rating to an item based on observed ratings from the population. As with all observational studies, hidden confounders, which are factors that affect both item exposures and user ratings, lead to a systematic bias in the estimation. Consequently, causal inference has been introduced in recommendations to address the influence of unobserved confounders. Observing that confounders in recommendations are usually shared among items and are therefore multi-cause confounders, we model the recommendation as a multi-cause multi-outcome (MCMO) inference problem. Specifically, to remedy the confounding bias, we estimate user-specific latent variables that render the item exposures independent Bernoulli trials. The generative distribution is parameterized by a DNN with factorized logistic likelihood and the intractable posteriors are estimated by variational inference. Controlling these factors as substitute confounders, under mild assumptions, can eliminate the bias incurred by multi-cause confounders. Furthermore, we show that MCMO modeling may lead to high variance due to scarce observations associated with the high-dimensional treatment space. Therefore, we theoretically demonstrate that controlling user features as pre-treatment variables can substantially improve sample efficiency and alleviate overfitting. Empirical studies on both simulated and real-world datasets demonstrate that the proposed deep causal recommender shows more robustness to unobserved confounders than state-of-the-art causal recommenders. Codes and datasets are released at https://github.com/yaochenzhu/Deep-Deconf.