Biblio

Keystroke dynamics study the way in which users input text via their keyboards, which is unique to each individual, and can form a component of a behavioral biometric system to improve existing account security. Keystroke dynamics systems on free-text data use n-graphs that measure the timing between consecutive keystrokes to distinguish between users. Many algorithms require 500, 1,000, or more keystrokes to achieve EERs of below 10%. In this paper, we propose an instance-based graph comparison algorithm to reduce the number of keystrokes required to authenticate users. Commonly used features such as monographs and digraphs are investigated. Feature importance is determined and used to construct a fused classifier. Detection error tradeoff (DET) curves are produced with different numbers of keystrokes. The fused classifier outperforms the state-of-the-art with EERs of 7.9%, 5.7%, 3.4%, and 2.7% for test samples of 50, 100, 200, and 500 keystrokes.

Research on keystroke dynamics has the good potential to offer continuous authentication that complements conventional authentication methods in combating insider threats and identity theft before more harm can be done to the genuine users. Unfortunately, the large amount of data required by free-text keystroke authentication often contain personally identifiable information, or PII, and personally sensitive information, such as a user's first name and last name, username and password for an account, bank card numbers, and social security numbers. As a result, there are privacy risks associated with keystroke data that must be mitigated before they are shared with other researchers. We conduct a systematic study to remove PII's from a recent large keystroke dataset. We find substantial amounts of PII's from the dataset, including names, usernames and passwords, social security numbers, and bank card numbers, which, if leaked, may lead to various harms to the user, including personal embarrassment, blackmails, financial loss, and identity theft. We thoroughly evaluate the effectiveness of our detection program for each kind of PII. We demonstrate that our PII detection program can achieve near perfect recall at the expense of losing some useful information (lower precision). Finally, we demonstrate that the removal of PII's from the original dataset has only negligible impact on the detection error tradeoff of the free-text authentication algorithm by Gunetti and Picardi. We hope that this experience report will be useful in informing the design of privacy removal in future keystroke dynamics based user authentication systems.

Recent insights on language and vision with neural networks have been successfully applied to simple single-image visual question answering. However, to tackle real-life question answering problems on multimedia collections such as personal photos, we have to look at whole collections with sequences of photos or videos. When answering questions from a large collection, a natural problem is to identify snippets to support the answer. In this paper, we describe a novel neural network called Focal Visual-Text Attention network (FVTA) for collective reasoning in visual question answering, where both visual and text sequence information such as images and text metadata are presented. FVTA introduces an end-to-end approach that makes use of a hierarchical process to dynamically determine what media and what time to focus on in the sequential data to answer the question. FVTA can not only answer the questions well but also provides the justifications which the system results are based upon to get the answers. FVTA achieves state-of-the-art performance on the MemexQA dataset and competitive results on the MovieQA dataset.

Text-based CAPTCHAs are still commonly used to attempt to prevent automated access to web services. By displaying an image of distorted text, they attempt to create a challenge image that OCR software can not interpret correctly, but a human user can easily determine the correct response to. This work focuses on a CAPTCHA used by a popular Chinese language question-and-answer website and how resilient it is to modern machine learning methods. While the majority of text-based CAPTCHAs focus on transcription tasks, the CAPTCHA solved in this work is based on localization of inverted symbols in a distorted image. A convolutional neural network (CNN) was created to evaluate the likelihood of a region in the image belonging to an inverted character. It is used with a feature map and clustering to identify potential locations of inverted characters. Training of the CNN was performed using curriculum learning and compared to other potential training methods. The proposed method was able to determine the correct response in 95.2% of cases of a simulated CAPTCHA and 67.6% on a set of real CAPTCHAs. Potential methods to increase difficulty of the CAPTCHA and the success rate of the automated solver are considered.

As a multimedia security mechanism, CAPTCHAs are completely automated public turing test to tell computers and humans apart. Although cracking CAPTCHA has been explored for many years, it is still a challenging problem for real practice. In this demo, we present a text based CAPTCHA cracking system by using convolutional neural networks(CNN). To solve small sample problem, we propose to combine conditional deep convolutional generative adversarial networks(cDCGAN) and CNN, which makes a tremendous progress in accuracy. In addition, we also select multiple models with low pearson correlation coefficients for majority voting ensemble, which further improves the accuracy. The experimental results show that the system has great advantages and provides a new mean for cracking CAPTCHAs.

At a time when all it takes to open a Twitter account is a mobile phone, the act of authenticating information encountered on social media becomes very complex, especially when we lack measures to verify digital identities in the first place. Because the platform supports anonymity, fake news generated by dubious sources have been observed to travel much faster and farther than real news. Hence, we need valid measures to identify authors of misinformation to avert these consequences. Researchers propose different authorship attribution techniques to approach this kind of problem. However, because tweets are made up of only 280 characters, finding a suitable authorship attribution technique is a challenge. This research aims to classify authors of tweets by comparing machine learning methods like logistic regression and naive Bayes. The processes of this application are fetching of tweets, pre-processing, feature extraction, and developing a machine learning model for classification. This paper illustrates the text classification for authorship process using machine learning techniques. In total, there were 46,895 tweets used as both training and testing data, and unique features specific to Twitter were extracted. Several steps were done in the pre-processing phase, including removal of short texts, removal of stop-words and punctuations, tokenizing and stemming of texts as well. This approach transforms the pre-processed data into a set of feature vector in Python. Logistic regression and naive Bayes algorithms were applied to the set of feature vectors for the training and testing of the classifier. The logistic regression based classifier gave the highest accuracy of 91.1% compared to the naive Bayes classifier with 89.8%.

SMS spam refers to undesired text message. Machine Learning methods for anti-spam filters have been noticeably effective in categorizing spam messages. Dataset used in this research is known as Tiago's dataset. Crucial step in the experiment was data preprocessing, which involved reducing text to lower case, tokenization, removing stopwords. Convolutional Neural Network was the proposed method for classification. Overall model's accuracy was 98.4%. Obtained model can be used as a tool in many applications.

Spam emails have been a chronic issue in computer security. They are very costly economically and extremely dangerous for computers and networks. Despite of the emergence of social networks and other Internet based information exchange venues, dependence on email communication has increased over the years and this dependence has resulted in an urgent need to improve spam filters. Although many spam filters have been created to help prevent these spam emails from entering a user's inbox, there is a lack or research focusing on text modifications. Currently, Naive Bayes is one of the most popular methods of spam classification because of its simplicity and efficiency. Naive Bayes is also very accurate; however, it is unable to correctly classify emails when they contain leetspeak or diacritics. Thus, in this proposes, we implemented a novel algorithm for enhancing the accuracy of the Naive Bayes Spam Filter so that it can detect text modifications and correctly classify the email as spam or ham. Our Python algorithm combines semantic based, keyword based, and machine learning algorithms to increase the accuracy of Naive Bayes compared to Spamassassin by over two hundred percent. Additionally, we have discovered a relationship between the length of the email and the spam score, indicating that Bayesian Poisoning, a controversial topic, is actually a real phenomenon and utilized by spammers.

This paper presents an assessment of continuous verification using linguistic style as a cognitive biometric. In stylometry, it is widely known that linguistic style is highly characteristic of authorship using representations that capture authorial style at character, lexical, syntactic, and semantic levels. In this work, we provide a contrast to previous efforts by implementing a one-class classification problem using Isolation Forests. Our approach demonstrates the usefulness of this classifier for accurately verifying the genuine user, and yields recognition accuracy exceeding 98% using very small training samples of 50 and 100-character blocks.

The paper dwells on the peculiarities of stylometry technologies usage to determine the style of the author publications. Statistical linguistic analysis of the author's text allows taking advantage of text content monitoring based on Porter stemmer and NLP methods to determine the set of stop words. The latter is used in the methods of stylometry to determine the ownership of the analyzed text to a specific author in percentage points. There is proposed a formal approach to the definition of the author's style of the Ukrainian text in the article. The experimental results of the proposed method for determining the ownership of the analyzed text to a particular author upon the availability of the reference text fragment are obtained. The study was conducted on the basis of the Ukrainian scientific texts of a technical area.

Although various techniques have been proposed to generate adversarial samples for white-box attacks on text, little attention has been paid to a black-box attack, which is a more realistic scenario. In this paper, we present a novel algorithm, DeepWordBug, to effectively generate small text perturbations in a black-box setting that forces a deep-learning classifier to misclassify a text input. We develop novel scoring strategies to find the most important words to modify such that the deep classifier makes a wrong prediction. Simple character-level transformations are applied to the highest-ranked words in order to minimize the edit distance of the perturbation. We evaluated DeepWordBug on two real-world text datasets: Enron spam emails and IMDB movie reviews. Our experimental results indicate that DeepWordBug can reduce the classification accuracy from 99% to 40% on Enron and from 87% to 26% on IMDB. Our results strongly demonstrate that the generated adversarial sequences from a deep-learning model can similarly evade other deep models.

Among many research efforts devoted to automated art investigations, the problem of quantification of literary style remains current. Meanwhile, linguists and computer scientists have tried to sort out texts according to their types or authors. We use the recently-introduced p-leader multifractal formalism to analyze a corpus of novels written for adults and young adults, with the goal of assessing if a difference in style can be found. Our results agree with the interpretation that novels written for young adults largely follow conventions of the genre, whereas novels written for adults are less homogeneous.

The notion of style is pivotal to literature. The choice of a certain writing style moulds and enhances the overall character of a book. Stylometry uses statistical methods to analyze literary style. This work aims to build a recommendation system based on the similarity in stylometric cues of various authors. The problem at hand is in close proximity to the author attribution problem. It follows a supervised approach with an initial corpus of books labelled with their respective authors as training set and generate recommendations based on the misclassified books. Results in book similarity are substantiated by domain experts.

Spammers use automated content spinning techniques to evade plagiarism detection by search engines. Text spinners help spammers in evading plagiarism detectors by automatically restructuring sentences and replacing words or phrases with their synonyms. Prior work on spun content detection relies on the knowledge about the dictionary used by the text spinning software. In this work, we propose an approach to detect spun content and its seed without needing the text spinner's dictionary. Our key idea is that text spinners introduce stylometric artifacts that can be leveraged for detecting spun documents. We implement and evaluate our proposed approach on a corpus of spun documents that are generated using a popular text spinning software. The results show that our approach can not only accurately detect whether a document is spun but also identify its source (or seed) document - all without needing the dictionary used by the text spinner.

For mobile phone users, short message service (SMS) is the most commonly used text-based communication type on mobile devices. Users can interact with other users and services via SMS. For example, users can send private messages, use information services, apply for a job advertisement, conduct bank transactions, and so on. Users should be very careful when using SMS. During the sending of SMS, the message content should be aware that it can be captured and act accordingly. Based on these findings, the elderly, called as “Silent Generation” which represents 70 years or older adults, are text messaging much more than they did in the past. Therefore, they need solutions which are both simple and secure enough if there is a need to send sensitive information via SMS. In this study, we propose and develop an android application to secure text messages. The application has a simple and easy-to-use graphical user interface but provides significant security.

In recent years, with the advances in JavaScript engines and the adoption of HTML5 APIs, web applications begin to show a tendency to shift their functionality from the server side towards the client side, resulting in dense and complex interactions with HTML documents using the Document Object Model (DOM). As a consequence, client-side vulnerabilities become more and more prevalent. In this paper, we focus on DOM-sourced Cross-site Scripting (XSS), which is a kind of severe but not well-studied vulnerability appearing in browser extensions. Comparing with conventional DOM-based XSS, a new attack surface is introduced by DOM-sourced XSS where the DOM could become a vulnerable source as well besides common sources such as URLs and form inputs. To discover such vulnerability, we propose a detecting framework employing hybrid analysis with two phases. The first phase is the lightweight static analysis consisting of a text filter and an abstract syntax tree parser, which produces potential vulnerable candidates. The second phase is the dynamic symbolic execution with an additional component named shadow DOM, generating a document as a proof-of-concept exploit. In our large-scale real-world experiment, 58 previously unknown DOM-sourced XSS vulnerabilities were discovered in user scripts of the popular browser extension Greasemonkey.

Social media has become an important platform for people to express opinions, share information and communicate with others. Detecting and tracking topics from social media can help people grasp essential information and facilitate many security-related applications. As social media texts are usually short, traditional topic evolution models built based on LDA or HDP often suffer from the data sparsity problem. Recently proposed topic evolution models are more suitable for short texts, but they need to manually specify topic number which is fixed during different time period. To address these issues, in this paper, we propose a nonparametric topic evolution model for social media short texts. We first propose the recurrent semantic dependent Chinese restaurant process (rsdCRP), which is a nonparametric process incorporating word embeddings to capture semantic similarity information. Then we combine rsdCRP with word co-occurrence modeling and build our short-text oriented topic evolution model sdTEM. We carry out experimental studies on Twitter dataset. The results demonstrate the effectiveness of our method to monitor social media topic evolution compared to the baseline methods.

Guidelines, directives, and policy statements are usually presented in ``linear'' text form - word after word, page after page. However necessary, this practice impedes full understanding, obscures feedback dynamics, hides mutual dependencies and cascading effects and the like, - even when augmented with tables and diagrams. The net result is often a checklist response as an end in itself. All this creates barriers to intended realization of guidelines and undermines potential effectiveness. We present a solution strategy using text as ``data'', transforming text into a structured model, and generate a network views of the text(s), that we then can use for vulnerability mapping, risk assessments and control point analysis. We apply this approach using two NIST reports on cybersecurity of smart grid, more than 600 pages of text. Here we provide a synopsis of approach, methods, and tools. (Elsewhere we consider (a) system-wide level, (b) aviation e-landscape, (c) electric vehicles, and (d) SCADA for smart grid).

Signature extraction is a critical preprocessing step in forensic log analysis because it enables sophisticated analysis techniques to be applied to logs. Currently, most signature extraction frameworks either use rule-based approaches or handcrafted algorithms. Rule-based systems are error-prone and require high maintenance effort. Hand-crafted algorithms use heuristics and tend to work well only for specialized use cases. In this paper we present a novel approach to extract signatures from forensic logs that is based on a neural language model. This language model learns to identify mutable and non-mutable parts in a log message. We use this information to extract signatures. Neural language models have shown to work extremely well for learning complex relationships in natural language text. We experimentally demonstrate that our model can detect which parts are mutable with an accuracy of 86.4%. We also show how extracted signatures can be used for clustering log lines.

This article describes a framework for semantic annotation of texts that are submitted for forensic analysis, based on Frame Semantics, and a knowledge base of Forensic Frames - FrameFOR. We demonstrate through experimental evaluations that the application of the Semantic Role Labeling (SRL) techniques and Natural Language Processing (NLP) in digital forensic increases the performance of the forensic experts in terms of agility, precision and recall.

In this paper we present results of a research on automatic extremist text detection. For this purpose an experimental dataset in the Russian language was created. According to the Russian legislation we cannot make it publicly available. We compared various classification methods (multinomial naive Bayes, logistic regression, linear SVM, random forest, and gradient boosting) and evaluated the contribution of differentiating features (lexical, semantic and psycholinguistic) to classification quality. The results of experiments show that psycholinguistic and semantic features are promising for extremist text detection.

Data analytics is being increasingly used in cyber-security problems, and found to be useful in cases where data volumes and heterogeneity make it cumbersome for manual assessment by security experts. In practical cyber-security scenarios involving data-driven analytics, obtaining data with annotations (i.e. ground-truth labels) is a challenging and known limiting factor for many supervised security analytics task. Significant portions of the large datasets typically remain unlabelled, as the task of annotation is extensively manual and requires a huge amount of expert intervention. In this paper, we propose an effective active learning approach that can efficiently address this limitation in a practical cyber-security problem of Phishing categorization, whereby we use a human-machine collaborative approach to design a semi-supervised solution. An initial classifier is learnt on a small amount of the annotated data which in an iterative manner, is then gradually updated by shortlisting only relevant samples from the large pool of unlabelled data that are most likely to influence the classifier performance fast. Prioritized Active Learning shows a significant promise to achieve faster convergence in terms of the classification performance in a batch learning framework, and thus requiring even lesser effort for human annotation. An useful feature weight update technique combined with active learning shows promising classification performance for categorizing Phishing/malicious URLs without requiring a large amount of annotated training samples to be available during training. In experiments with several collections of PhishMonger's Targeted Brand dataset, the proposed method shows significant improvement over the baseline by as much as 12%.

The U.S Government has been the target for cyberattacks from all over the world. Just recently, former President Obama accused the Russian government of the leaking emails to Wikileaks and declared that the U.S. might be forced to respond. While Russia denied involvement, it is clear that the U.S. has to take some defensive measures to protect its data infrastructure. Insider threats have been the cause of other sensitive information leaks too, including the infamous Edward Snowden incident. Most of the recent leaks were in the form of text. Due to the nature of text data, security classifications are assigned manually. In an adversarial environment, insiders can leak texts through E-mail, printers, or any untrusted channels. The optimal defense is to automatically detect the unstructured text security class and enforce the appropriate protection mechanism without degrading services or daily tasks. Unfortunately, existing Data Leak Prevention (DLP) systems are not well suited for detecting unstructured texts. In this paper, we compare two recent approaches in the literature for text security classification, evaluating them on actual sensitive text data from the WikiLeaks dataset.

Statistical prediction models can be an effective technique to identify vulnerable components in large software projects. Two aspects of vulnerability prediction models have a profound impact on their performance: 1) the features (i.e., the characteristics of the software) that are used as predictors and 2) the way those features are used in the setup of the statistical learning machinery. In a previous work, we compared models based on two different types of features: software metrics and term frequencies (text mining features). In this paper, we broaden the set of models we compare by investigating an array of techniques for the manipulation of said features. These techniques fall under the umbrella of dimensionality reduction and have the potential to improve the ability of a prediction model to localize vulnerabilities. We explore the role of dimensionality reduction through a series of cross-validation and cross-project prediction experiments. Our results show that in the case of software metrics, a dimensionality reduction technique based on confirmatory factor analysis provided an advantage when performing cross-project prediction, yielding the best F-measure for the predictions in five out of six cases. In the case of text mining, feature selection can make the prediction computationally faster, but no dimensionality reduction technique provided any other notable advantage.

The chaotic system and cryptography have some common features. Due to the close relationship between chaotic system and cryptosystem, researchers try to combine the chaotic system with cryptosystem. In this study, security analysis of an encryption algorithm which aims to encrypt the data with ECG signals and chaotic functions was performed using the Logistic map in text encryption and Henon map in image encryption. In the proposed algorithm, text and image data can be encrypted at the same time. In addition, ECG signals are used to determine the initial conditions and control parameters of the chaotic functions used in the algorithm to personalize of the encryption algorithm. In this cryptanalysis study, the inadequacy of the mentioned process and the weaknesses of the proposed method have been determined. Encryption algorithm has not sufficient capacity to provide necessary security level of key space and secret key can be obtained with only one plaintext/ciphertext pair with chosen-plaintext attack.