User Identification Problem Research Articles

User identification across online social networks based on gated multi-feature extraction

User identification is an essential technical support for downstream tasks such as recommendation systems, information retrieval, and collaborative filtering. Computing the similarity between user display names through classifiers is an effective solution for user identification across social networks. However, there are two problems with existing methods. Applying expert domain knowledge to extract handcrafted features of display names ignores the semantic information, resulting in poor performance of these methods. Selecting influential display names, and handcrafted features in user identification problems are also one of the difficulties. To solve these two problems, we propose a method based on the multi-feature fusion of display names using gated units. First, we extract the deep semantic features of display names through the BERT pre-trained multi-language model. Then, the gated mechanism is applied to select the handcrafted features we extracted to retain the essential features. Then, the adaptive factors are used to fuse handcrafted features and deep features to obtain user identification results across social networks. Finally, the efficiency of our model is verified on three constructed real-world multilingual display names datasets across multiple online social networks and compared with existing state-of-the-art methods. Experimental results show that the proposed algorithm outperforms the compared algorithms.

SPAMMER DETECTION AND FAKE USER IDENTIFICATION ON SOCIAL NETWORKS

The project titled "Spammer Detection and Fake User Identification on Social Networks" aims to explore and implement a novel approach to concealing information within digital images while preserving their visual integrity. With the exponential growth of social media platforms like Instagram, the issue of spam and fake user accounts has become increasingly prevalent, posing significant challenges to the integrity and user experience of these platforms. In this project, a comprehensive approach to address the problem of spammer detection and fake user identification on Instagram is proposed. Leveraging machine learning techniques, including Decision Tree Classifier, KNN Classifier, Random Forest Classifier, and Logistic Regression Algorithm, the project aims to develop robust models capable of automatically identifying and flagging suspicious accounts. By analyzing various features such as user behaviour patterns, engagement metrics, and content characteristics, these classifiers will be trained to differentiate between genuine and fake accounts effectively. The project's ultimate goal is to contribute to the enhancement of Instagram's spam detection mechanisms, fostering a safer and more authentic social media environment for users. Key Words: Spam detection, Fake user identification, KNN Classifier, Decision Tree Classifier, Random Forest Classifier, Logistic Regression Algorithm.

User Identification Across Multiple Social Networks Based on Naive Bayes Model.

Recently, the problem of user identification across multiple social networks (UIAMSNs) has attracted considerable attention since it is a prerequisite for many downstream tasks and applications. Although substantial network feature-based approaches have been proposed to solve the UIAMSNs' problem, the matching degree in most of the current works is given by experience, which lacks a solid theoretical basis. To alleviate the above predicament, we propose a user identification algorithm based on naive Bayes model (UI-NBM) within the network feature-based framework. First, a matching degree index is designed based on the naive Bayes model, which can accurately measure the contributions of different common matched node pairs (MNPs) to the connection probability of unmatched node pairs (UMNPs). Second, the matching degrees of all UMNPs are formulated as the product of matrices, giving rise to the great reduction of the time complexity and the compact expression; Finally, with the idea of recursion process, more UMNPs can be iteratively predicted even when only a small amount of prior information (i.e., a few number of MNPs) is known. The experimental results on the synthetic and real cross platforms demonstrate that the method outperforms the baseline methods within the feature-based framework.

A Binary-Search-Based Locality-Sensitive Hashing Method for Cross-Site User Identification

In recent years, we have witnessed increasing attention paid to the problem of cross-site user identification (CSUI) with the bloom of social media. Despite noticeable progress in this field, the problem of enormous computation posed by the full-scale pairwise comparison still remains unsolved, which constrains the real application of the state-of-the-art approaches, especially when the user number reaches tens of millions. To address this issue, we propose a novel trajectory-oriented method, which aims at incorporating the locality-sensitive hashing (LSH) technology into the strategy of approximate nearest neighbors searching. Specifically, we design an LSH function based on binary search (BLSH) for processing user trajectory easily, and by searching the BLSH buckets, we cluster similar users to avoid the onerous full-scale pairwise comparison. Moreover, a hierarchy of discrete attention is devised to further improve the effectiveness, which develops our method into the mature one, hierarchically attentioned binary-search-based LSH (HA-BLSH). Extensive experiments on ground-truth datasets demonstrate that our HA-BLSH outperforms the cutting-edge approaches on efficiency and effectiveness. The further discussion indicates that our method is also quite flexible and able to accommodate more scenarios.

Let IoT Know You Better: User Identification and Emotion Recognition Through Millimeter-Wave Sensing

Emotion recognition, particularly contactless recognition via wireless sensing, has shown its promise in diverse applications. However, the previous works only focus on emotions rather than the person, i.e., the premise is already knowing who the subject is, without considering the issue of identifying subjects. We envision that user identification and emotion recognition together will bring more adaptive and personalized Internet of Things applications, e.g., a smart home system can react to specific emotions of a specific user, independently. In this work, we move forward to investigate the problem of simultaneous user identification, using only physiological indicators embedded in wireless signals reflected off from targets. Toward the objective, in this article, we first carry out a comprehensive measurement study, which validates the feasibility of simultaneous user identification and emotion recognition. Moreover, the measurement also discovers that the key challenge lies in the limitation of artificial features and the substantial emotion feature deviation across different days, which hinders accurate and robust sensing. To resolve the challenge, we design two multiscale neural networks, incorporated with a custom-built feature attention mechanism, so as to obtain rich feature expression and, thus, enhance the important features for accurate recognition. We prototype mmEMO using a commercial off-the-shelf millimeter-wave radar and experimental evaluation shows that mmEMO can achieve 87.68% user identification accuracy and 80.59% emotion recognition accuracy, respectively.

User recognition in participatory sensing systems using deep learning based on spectro-temporal representation of accelerometer signals

Identification of the actual sensor data provider is important in smartphone-based participatory sensing systems. More specifically, the participatory sensing systems must be able to identify the actual users from smartphone-embedded sensors’ data in a fast, silent, and unobtrusive manner, ensuring that appropriate privacy protection and authentication are provided automatically. In this context, this paper investigates smartphone-equipped accelerometer signal data to determine a person’s distinctive gait, allowing different providers to be recognized. The problem of user identification is formulated as an image classification task by changing the accelerometer signal into a 2D spectrotemporal image representation by calculating Short-Term Fourier Transform (STFT) of the signal. A deep learning based custom CNN model is proposed and evaluated on real-world benchmark dataset (containing 60 million accelerometer data samples of 387 users) for user identification. The efficacy of the proposed deep learning based approach is verified by comparing its performance with several baseline methods including Multi-layer Perceptron, K-Nearest Neighbours, Logistic Regression, Random Forest, and Decision Tree. The results of the experiments confirm that the proposed technique yields better performance (accuracy = 0.98, precision = 0.96, recall = 0.98, F1-score = 0.96, AUC = .9995) with respect to the baselines. Furthermore, the proposed model outperforms other CNN architectures (e.g., Lenet-5, AlexNet, and ZFNet) across all performance metrics.

COMPUTER USER IDENTIFICATION POSSIBILITIES FOLLOWING THE KEYPAD DISPLAYS

Identification of a computer user in modern conditions allows not only to ensure the protection of information, but also has a special forensic significance. The developed systems of user authentication by keyboard handwriting, based on the identification of an individual dynamic stereotype of working with the keyboard (the intervals between keystrokes and the time of holding the keys), make it possible to identify the user only during data entry. When investigating crimes using computer technology, in most cases it is necessary to install the user after a considerable time after using the computer. The conducted pilot study found that the individual features of the dynamic stereotype of working with the keyboard are manifested in stable traces-displays on the keyboard keys. The features of the stereotype are manifested in the area where the keys are touched. At the same time, the touch area can be different for different keys: top, bottom, right, left, center, one of the corners. This creates a high variability, individuality of working with the keyboard. With prolonged use, micro-damages of the surface are formed in these areas - scuffs, scratches. With relatively short-term use, the areas with the greatest stratification of the fatty substance can be used as traces. To solve the problem of user identification, it is necessary to solve the problem of distinguishing the traces of several users on the same keyboard.

Voice Recognition Security Reliability Analysis Using Deep Learning Convolutional Neural Network Algorithm

This study discusses the reliability analysis of voice recognition security using the deep learning convolutional neural network (CNN) algorithm. The CNN algorithm has learning advantages in that it is safer, faster, and more accurate. CNN also can solve user identification problems in large amounts of data. The measured voice input is ten types of user's voice with the number of iterations of 6000, 12000, and 15000 sound files. Furthermore, voice extraction features are performed to recognize conversations and retain information that is very much needed. After that, the voice file iteration data is trained to register the user's voice so that a trained model is obtained. These results measure performance (confusion matrix) to analyze the actual value compared to the predicted value in the CNN algorithm. The results obtained are that the best accuracy is obtained at 15000 sound file iterations, 96.87%, 12000 sound file iterations get 96.30%, and 6000 sound file iterations get 95.77%. CNN's performance data shows that 15000 iterations of voice files produce high accuracy. Voice recognition security helps provide high security and maintain the privacy of one's identity.

Некоторые подходы к оценке информативности параметров идентификации пользователя по клавиатурному почерку на основе поведенческой биометрии

Domestic normative documents, in particular the National Standard of the Russian Federation on Information Protection, outlines three levels of confidence in the results of identification, based on some degree of confidence in the authenticity of the subject of access, but not containing specific values. The article proposes approaches to qualitative assessment of informativeness of user identification parameters by keyboard handwriting on the basis of behavioral biometrics. Objective of the study. Evaluation of informativeness of user identification parameters by keyboard handwriting. Materials and methods of research. To determine approaches to assessing the informativeness of parameters, we studied and analyzed known scientific results and practical solutions on the problem of user identification in information and telecommunication computer systems, available in various open sources in Russian and English. Main results. We identified the currently relevant set of parameters, identifying the user by the keyboard handwriting; the list of noises affecting the informativeness of the parameters. Average values of identification reliability levels were calculated. The basic and additional criteria for evaluating the informativeness of the parameters were defined. False access and false denial rates and overall system evaluation were selected as basic criteria for evaluating reliability of identification by handwriting. As additional criteria the speed of operation, ease of use, the cost of the system are highlighted. For the main criteria the indicators that we propose to use as evaluation characteristics, which could be used to judge the degree of achievement of the criterion: the stability of the parameter in distinguishing users from each other; the number of implementations of the parameter, required to ensure its stability in distinguishing users. The qualitative characteristics of the degree of informativeness of the parameters on the main criteria are given. The directions for further research are defined. Conclusion. The list of parameters and noises presented in the article, the assessment of identification reliability by criteria and corresponding indicators will be useful to developers and researchers for further refinement of identification methods to improve their reliability.

ИСПОЛЬЗОВАНИЕ ЛАТЕНТНО-СЕМАНТИЧЕСКОГО АНАЛИЗА ПРИ ПОДГОТОВКЕ ДАННЫХ ДЛЯ ИДЕНТИФИКАЦИИ АНОНИМНЫХ ПОЛЬЗОВАТЕЛЕЙ ПО ЦИФРОВЫМ ОТПЕЧАТКАМ

Digital fingerprints changings over time as a result of system, plugins, browsers, installation of various programs updates and fonts is a serious problem in the method of tracking and identifying users using a browser (Fingerprinting (FP) of a web browser). The set of parsed attributes can contain both metrical and categorical (mostly non-numeric) values, for example, parameters such as user-agent, webgl, canvas, etc. Considering this, it is required to pre-encode them for the convenience of further processing. For these purposes, artificial intelligence technologies, including the processing of text in natural languages NLP (Natural Language Processing), are widely used. The aim of the research is to analyze the peculiarities of the implementation of latent-semantic analysis (LSA) in the preparation and analysis of FP data for the identification of anonymous users. Methods. A comparative analysis of the common ways of converting categorical values of fingerprint attributes (FP) into numeric One-Hot-Encoding, Label-Encoder, LSA for identifying anonymous users with a predetermined number of possible values of categorical features is carried out. Results. The advantage of the LSA algorithm over One-Hot-Encoding, Label-Encoder is shown. The possibility of clustering implementation within the framework of the user identification problem by visualizing FP (FP) relative to hidden semantic topics using the LSA model of latent semantic analysis is shown. It is shown that with a small number of hid& den topics using the obtained vectors of objects and vectors of terms for assessing the similarity of two FPs, the proposed model allows us to confidently classify the input FP to a common topic. With the help of the obtained vectors of objects and vectors of terms for assessing the similarity of two FP objects, it becomes possible to apply various measures of cluster proximity: Euclidean distance, cosine measure, etc.

Development of security systems using DNN and i & x-vector classifiers

The widespread use of biometric systems entails increased interest from cybercriminals aimed at developing attacks to crack them. Thus, the development of biometric identification systems must be carried out taking into account protection against these attacks. The development of new methods and algorithms for identification based on the presentation of randomly generated key features from the biometric base of user standards will help to minimize the disadvantages of the above methods of biometric identification of users. We present an implementation of a security system based on voice identification as an access control key and a verification algorithm developed using MATLAB function blocks that can authenticate a person's identity by his or her voice. Our research has shown an accuracy of 90 % for this user identification system for individual voice characteristics. It has been experimentally proven that traditional MFCCs using DNN and i and x-vector classifiers can achieve good results. The paper considers and analyzes the most well-known approaches from the literature to the problem of user identification by voice: dynamic programming methods, vector quantization, mixtures of Gaussian processes, hidden Markov model. The developed software package for biometric identification of users by voice and the method of forming the user's voice standards implemented in the complex allows reducing the number of errors in identifying users of information systems by voice by an average of 1.5 times. Our proposed system better defines voice recognition in terms of accuracy, security and complexity. The application of the results obtained will improve the security of the identification process in information systems from various attacks.

Influence of destabilizing factors on the stability of user's handwritten signature indicators

Consideration is given to the question of user`s handwritten signature parameters informativeness and stability during authentication. An identifier that uses biometric characteristics is inextricably linked to the user and it is almost impossible to use it without authorization. It is proposed to use dynamic biometric characteristics of users. Their advantage is that due to the presence of a dynamic component, the probability of their forgery by an attacker is very low. A handwritten signature is used as a biometric characteristic of the user. A handwritten signature is a socially and legally recognized biometric characteristic used for human authentication. It has a rather complex structure and high detail - all this makes solving the problem of user identification by mathematical methods quite complex and requires high computational costs. Another significant disadvantage is that handwritten authentication systems require the installation of additional specialized equipment, which makes the use of such systems as an ordinary means of authentication very expensive. Nowadays the presence of mobile devices in almost all users has made it possible to form the idea of using them in authentication systems. Thanks to that a scheme for implementing a computer security system against unauthorized access based on handwritten signatures using Android-based mobile devices as signature input devices were proposed. An algorithm based on Heming's distance was chosen to implement user tolerance. According to the selected algorithm, a method for forming a biometric vector has been developed. The optimal characteristics are investigated and the efficiency of using the proposed form biometric characteristics vector is estimated. The speed of movement at certain intervals and the inclination angle of the vector interval were chosen as indicators of the handwritten signature. It is offered to estimate stability in time and dependence of the chosen biometric signs on the following factors: emotional and physical condition of the user, and also time of day at the moment of authentication. Developed a software application for the Android operating system, which collects the time characteristics and values of the proposed factors for time characteristics vectors, as well as allows to export the accumulated data.

BIM modeling in construction taking into account data analysis and user identification problems

BIM modeling in construction taking into account data analysis and user identification problems

EFFICIENCY OF REMOTE TECHNOLOGIES ON THE APPROACHING THE CONTENT OF A E-LEARNING MATHEMATICS COURSE USING MOODLE SYSTEM: CASE STUDY

In the higher education system, distance learning using electronic educational courses is becoming the most relevant and widely demanded way of learning. This is primarily due to the rapidly developing worldwide process of digitalization of education. Besides, in connection with the transition to a two-level education system, the number of hours in the curriculum devoted to students’ independent work has significantly increased. These technologies are a useful tool for organizing a student’s cognitive independence. The advantages are the ability to study at any convenient time, anywhere, the ability to learn at one’s own pace, and the ability to learn without interrupting work. On the other hand, there are some shortcomings, such as the problem of user identification and the need for the student to have strong motivation. This study aimed to show one of the options for organizing students' independent work on the example of the electronic educational course "Mathematics and Fundamentals of Mathematical Information Processing", created on the basis of the LMS MOODLE system. The experiment was attended by 30 students from two academic groups (15 people each) aged 16 to 17, studying at the Elabuga Institute of Kazan (Volga) Federal University. One of the groups mastered the discipline "Mathematics and Fundamentals of Mathematical Information Processing" in the traditional form, the other group - in the remote form using the electronic educational course developed by the authors. After studying the course in both groups, verification work and a survey were conducted. Processing the results of the control work using statistical methods made it possible to assess how much the success of students in mathematics depends on the choice of the form of training. The survey results allowed the authors to formulate ways to improve the quality of distance learning mathematics using an electronic educational course.

Selection of handwritten signature dynamic indicators for user authentication

Selection and evaluation of handwritten signature indicators in user authentication systems are considered. An important and unresolved issue of information security is the effective identification of the user who accesses confidential information. Traditional password protection has a number of disadvantages. As an alternative to the password system or its addition, user identification by biometric characteristics is considered. An identifier that uses biometric characteristics is inextricably linked to the user and it is almost impossible to use it without authorization. The decision about the truth of the user of biometric authentication systems is made on the basis of comparison of his template with the data entered when trying to authenticate. The template is formed on the basis of studying the selected individual characteristics of the user. To do this, we use the biometric characteristics of users, which reflect the dynamic, behavioral characteristics of the person. It is proposed to use a handwritten signature as a biometric characteristic of the user. A handwritten signature is a socially and legally recognized biometric characteristic used for human authentication. It has a rather complex structure and high detail - all this makes solving the problem of user identification by mathematical methods quite complex and requires high computational costs. Another significant disadvantage is that handwritten authentication systems require the installation of additional specialized equipment, which makes the use of such systems as an ordinary means of authentication very expensive. To solve this problem, a scheme for implementing a computer data protection system against unauthorized access based on handwritten signatures using mobile devices based on the Android operating system as signature input devices is proposed. A method of biometric vector generation for use in dynamic biometric user authentication systems is proposed. The optimal characteristics are investigated and the efficiency of using the proposed form biometric characteristics vector is estimated. Certificates of copyrights registration for software applications developed during the work were received.

Методы решения задач аутентификации и идентификации пользователя на основе анализа клавиатурного почерка

Nowadays, information technologies are used in almost every sphere of our life. Large amounts of personal and corporate data are stored in digital form. Thus, the problem of protecting this data from unauthorized access is raised. Access control subsystems and technologies used in these subsystems, in particular, methods of user authentication and identification, play an important role in this case. The paper provides an overview of methods for solving information system user authentication and identification problems based on the analysis of their keystroke dynamics. These tasks are considered in detail, a brief history of methods for keystroke dynamics authentication and identification is described, their advantages and disadvantages compared with other known methods are indicated and the relevance of this line of research is substantiated. The main stages of creating mathematical models for solving these problems are listed. The description of the methods used in each of these stages is also presented

Exploiting similarities of user friendship networks across social networks for user identification

User identification has been attracting considerable attention from academia. Due to the uniqueness and difficulty of faking friendship networks, some friendship-based methods have been presented to improve the identification performance. However, the information redundancies in k-hop (k > 1) neighbors and their contributions to user identification have not been fully analyzed in the existing work. Addressing these two issues helps to understand the problem of friendship-based user identification and to propose more effective solutions. In this paper, we first obtain ground-truth friendship networks across three popular social sites; then, we analyze the similarities of k-hop neighbors to fully characterize the information redundancies in the friendship network. We apply these information redundancies in several classifiers to study their contributions to user identification. Furthermore, we apply the friendship-based information redundancies jointly with the display-name-based information redundancies to perform user identification. The experiments show that (1) the similarities related to the 1-hop neighbors contribute to user identification much more than do the other similarities; (2) the information redundancies in the k-hop (k > 1) neighbors are also very useful for user identification; and (3) jointly applying display-name-based information redundancies can provide better performance and improve the universality of the identification method.

An unsupervised user identification algorithm using network embedding and scalable nearest neighbour

Most of the current studies on social network (SN) mainly focused on a single SN platform. Integration of SNs can provide more sufficient user behaviour data and more complete network structure, and thus is rewarding to an ocean of studies on social computing. Recognizing the identical users across SNs, or user identification, naturally bridges the SNs through users and has attracted extensive attentions. Due to the fragmentation, inconsistency and disruption of the accessible information among SNs, user identification is still an intractable problem. Different from the efforts implemented on user profiles and users’ content, many studies have noticed the accessibility and reliability of network structure in most of the SNs for addressing this issue. Although substantial achievements have been made, most of the current network structure-based solutions are supervised or semi-supervised and require some given identified users or seed users. In the scenarios where seed users are hard to obtain, it is laborious to label the seed users manually. In this study, we proposed an unsupervised scheme by employing the reliability and consistence of friend relationships in different SNs, termed Unsupervised Friend Relationship-based User Identification algorithm (UFRUI). The UFRUI first models the network structure and embeds the feature of each user into a vector using network embedding technique, and then converts the user identification problem into a nearest neighbour problem. Finally, the matching user is computed using the scalable nearest neighbour algorithm. Results of experiments demonstrated that UFRUI performs much better than current state-of-art network structure-based algorithm without seed users.

Matching user accounts based on user generated content across social networks

Matching user accounts can help us build better users’ profiles and benefit many applications. It has attracted much attention from both industry and academia. Most of existing works are mainly based on rich user profile attributes. However, in many cases, user profile attributes are unavailable, incomplete or unreliable, either due to the privacy settings or just because users decline to share their information. This makes the existing schemes quite fragile. Users often share their activities on different social networks. This provides an opportunity to overcome the above problem. We aim to address the problem of user identification based on User Generated Content (UGC). We first formulate the problem of user identification based on UGCs and then propose a UGC-based user identification model. A supervised machine learning based solution is presented. It has three steps: firstly, we propose several algorithms to measure the spatial similarity, temporal similarity and content similarity of two UGCs; secondly, we extract the spatial, temporal and content features to exploit these similarities; afterwards, we employ the machine learning method to match user accounts, and conduct the experiments on three ground truth datasets. The results show that the proposed method has given excellent performance with F1 values reaching 89.79%, 86.78% and 86.24% on three ground truth datasets, respectively. This work presents the possibility of matching user accounts with high accessible online data.

User identification across online social networks in practice: Pitfalls and solutions

To take advantage of the full range of services that online social networks (OSNs) offer, people commonly open several accounts on diverse OSNs where they leave lots of different types of profile information. The integration of these pieces of information from various sources can be achieved by identifying individuals across social networks. In this article, we address the problem of user identification by treating it as a classification task. Relying on common public attributes available through the official application programming interface (API) of social networks, we propose different methods for building negative instances that go beyond usual random selection so as to investigate the effectiveness of each method in training the classifier. Two test sets with different levels of discrimination are set up to evaluate the robustness of our different classifiers. The effectiveness of the approach is measured in real conditions by matching profiles gathered from Google+, Facebook and Twitter.