arXiv subjects

Christian Reuter

Publications and source records attributed to Christian Reuter.

16 recordsLinked to original sources

arXiv2026

The PIONEER Project: A PrIvacy companion for mOtivatioN and knowlEdge transfER

Remaining control over their private data is one of the key challenges in this century for users. We know from prior work that users are often neither in a position to fully grasp the content of the usually complicated texts, nor are they motivated to spend the time necessary to do so. We report on the progress made by the PIONEER project on a privacy support tool that combines knowledge transfer and persuasive elements to increase users' privacy awareness and motivation; thus empowering them to more privacy sovereignty. Throughout the research and design process, we consider user group specifics that may result in different requirements, e.g., for children, adolescents, parents, or elderly people. We further target sustainable behavior change by addressing different states of change, precisely: spark initial motivation, facilitate the creation of new habits, and encourage habituation of these habits in the long term (volition). Finally, we provide a privacy support tool demonstrator that can be utilized for research and education purposes, e.g., in school contexts.

cs.HC↗

arXiv2026

WOOTdroid: Whole-system Online On-device Tracing for Android

System auditing on Android faces two problems. First, existing syscall tracers lose events under load, silently overwriting entries faster than a user space reader can drain them. Second, security-relevant application behavior is mediated through Binder, Android's kernel IPC mechanism, and is therefore hidden from the syscall layer. The Binder parcels that the kernel does see carry no method names or typed arguments, a disconnect between low-level events and high-level behavior known as the semantic gap. Existing approaches address the semantic gap either by modifying the Android platform, making them difficult to adjust to OS updates, or by instrumenting the traced application in user space, which sophisticated adversaries can evade by bypassing the instrumented framework APIs. We present WOOTdroid, a design and prototype for on-device tracing on stock Android that addresses both problems without OS modification or application instrumentation. WDSys, an eBPF port of eAudit-style syscall auditing, runs on current Android with at most 3.6% Geekbench overhead and traces 33% more syscalls than ftrace. WDBind captures Binder parcels in the kernel and decodes them out-of-process against a framework signature table extracted via Java reflection. We demonstrate WOOTdroid on Pixel 9 devices running Android 16 with an end-to-end case study reconstructing ten security-relevant Binder transactions.

cs.CR↗

arXiv2025

Bandit on the Hunt: Dynamic Crawling for Cyber Threat Intelligence

Public information contains valuable Cyber Threat Intelligence (CTI) that is used to prevent future attacks. While standards exist for sharing this information, much appears in non-standardized news articles or blogs. Monitoring online sources for threats is time-consuming and source selection is uncertain. Current research focuses on extracting Indicators of Compromise from known sources, rarely addressing new source identification. This paper proposes a CTI-focused crawler using multi-armed bandit (MAB) and various crawling strategies. It employs SBERT to identify relevant documents while dynamically adapting its crawling path. Our system ThreatCrawl achieves a harvest rate exceeding 25% and expands its seed by over 300% while maintaining topical focus. Additionally, the crawler identifies previously unknown but highly relevant overview pages, datasets, and domains.

cs.CR↗

arXiv2025

A Survey of Machine Learning Models and Datasets for the Multi-label Classification of Textual Hate Speech in English

The dissemination of online hate speech can have serious negative consequences for individuals, online communities, and entire societies. This and the large volume of hateful online content prompted both practitioners', i.e., in content moderation or law enforcement, and researchers' interest in machine learning models to automatically classify instances of hate speech. Whereas most scientific works address hate speech classification as a binary task, practice often requires a differentiation into sub-types, e.g., according to target, severity, or legality, which may overlap for individual content. Hence, researchers created datasets and machine learning models that approach hate speech classification in textual data as a multi-label problem. This work presents the first systematic and comprehensive survey of scientific literature on this emerging research landscape in English (N=46). We contribute with a concise overview of 28 datasets suited for training multi-label classification models that reveals significant heterogeneity regarding label-set, size, meta-concept, annotation process, and inter-annotator agreement. Our analysis of 24 publications proposing suitable classification models further establishes inconsistency in evaluation and a preference for architectures based on Bidirectional Encoder Representation from Transformers (BERT) and Recurrent Neural Networks (RNNs). We identify imbalanced training data, reliance on crowdsourcing platforms, small and sparse datasets, and missing methodological alignment as critical open issues and formulate ten recommendations for research.

cs.CL↗

arXiv2024

ActiveLLM: Large Language Model-based Active Learning for Textual Few-Shot Scenarios

Active learning is designed to minimize annotation efforts by prioritizing instances that most enhance learning. However, many active learning strategies struggle with a `cold-start' problem, needing substantial initial data to be effective. This limitation reduces their utility in the increasingly relevant few-shot scenarios, where the instance selection has a substantial impact. To address this, we introduce ActiveLLM, a novel active learning approach that leverages Large Language Models such as GPT-4, o1, Llama 3, or Mistral Large for selecting instances. We demonstrate that ActiveLLM significantly enhances the classification performance of BERT classifiers in few-shot scenarios, outperforming traditional active learning methods as well as improving the few-shot learning methods ADAPET, PERFECT, and SetFit. Additionally, ActiveLLM can be extended to non-few-shot scenarios, allowing for iterative selections. In this way, ActiveLLM can even help other active learning strategies to overcome their cold-start problem. Our results suggest that ActiveLLM offers a promising solution for improving model performance across various learning setups.

cs.CL↗

arXiv2023

ExTRUST: Reducing Exploit Stockpiles with a Privacy-Preserving Depletion System for Inter-State Relationships

Cyberspace is a fragile construct threatened by malicious cyber operations of different actors, with vulnerabilities in IT hardware and software forming the basis for such activities, thus also posing a threat to global IT security. Advancements in the field of artificial intelligence accelerate this development, either with artificial intelligence enabled cyber weapons, automated cyber defense measures, or artificial intelligence-based threat and vulnerability detection. Especially state actors, with their long-term strategic security interests, often stockpile such knowledge of vulnerabilities and exploits to enable their military or intelligence service cyberspace operations. While treaties and regulations to limit these developments and to enhance global IT security by disclosing vulnerabilities are currently being discussed on the international level, these efforts are hindered by state concerns about the disclosure of unique knowledge and about giving up tactical advantages. This leads to a situation where multiple states are likely to stockpile at least some identical exploits, with technical measures to enable a depletion process for these stockpiles that preserve state secrecy interests and consider the special constraints of interacting states as well as the requirements within such environments being non-existent. This paper proposes such a privacy-preserving approach that allows multiple state parties to privately compare their stock of vulnerabilities and exploits to check for items that occur in multiple stockpiles without revealing them so that their disclosure can be considered. We call our system ExTRUST and show that it is scalable and can withstand several attack scenarios. Beyond the intergovernmental setting, ExTRUST can also be used for other zero-trust use cases, such as bug-bounty programs.

cs.CR↗

arXiv2023

ThreatCrawl: A BERT-based Focused Crawler for the Cybersecurity Domain

Publicly available information contains valuable information for Cyber Threat Intelligence (CTI). This can be used to prevent attacks that have already taken place on other systems. Ideally, only the initial attack succeeds and all subsequent ones are detected and stopped. But while there are different standards to exchange this information, a lot of it is shared in articles or blog posts in non-standardized ways. Manually scanning through multiple online portals and news pages to discover new threats and extracting them is a time-consuming task. To automize parts of this scanning process, multiple papers propose extractors that use Natural Language Processing (NLP) to extract Indicators of Compromise (IOCs) from documents. However, while this already solves the problem of extracting the information out of documents, the search for these documents is rarely considered. In this paper, a new focused crawler is proposed called ThreatCrawl, which uses Bidirectional Encoder Representations from Transformers (BERT)-based models to classify documents and adapt its crawling path dynamically. While ThreatCrawl has difficulties to classify the specific type of Open Source Intelligence (OSINT) named in texts, e.g., IOC content, it can successfully find relevant documents and modify its path accord ingly. It yields harvest rates of up to 52%, which are, to the best of our knowledge, better than the current state of the art. The results and source code will be made publicly available upon acceptance.

cs.CR↗

arXiv2023

The Landscape of User-centered Misinformation Interventions -- A Systematic Literature Review

Misinformation is one of the key challenges facing society today. User-centered misinformation interventions as digital countermeasures that exert a direct influence on users represent a promising means to deal with the large amounts of information available. While an extensive body of research on this topic exists, researchers are confronted with a diverse research landscape spanning multiple disciplines. This review systematizes the landscape of user-centered misinformation interventions to facilitate knowledge transfer, identify trends, and enable informed decision-making. Over 5,700 scholarly publications were screened and a systematic literature review (N=163) was conducted. A taxonomy was derived regarding intervention design (e.g., (binary) label), user interaction (active or passive), and timing (e.g., post exposure to misinformation). We provide a structured overview of approaches across multiple disciplines, and derive six overarching challenges for future research.

cs.HC↗

arXiv2022

CySecBERT: A Domain-Adapted Language Model for the Cybersecurity Domain

The field of cybersecurity is evolving fast. Experts need to be informed about past, current and - in the best case - upcoming threats, because attacks are becoming more advanced, targets bigger and systems more complex. As this cannot be addressed manually, cybersecurity experts need to rely on machine learning techniques. In the texutual domain, pre-trained language models like BERT have shown to be helpful, by providing a good baseline for further fine-tuning. However, due to the domain-knowledge and many technical terms in cybersecurity general language models might miss the gist of textual information, hence doing more harm than good. For this reason, we create a high-quality dataset and present a language model specifically tailored to the cybersecurity domain, which can serve as a basic building block for cybersecurity systems that deal with natural language. The model is compared with other models based on 15 different domain-dependent extrinsic and intrinsic tasks as well as general tasks from the SuperGLUE benchmark. On the one hand, the results of the intrinsic tasks show that our model improves the internal representation space of words compared to the other models. On the other hand, the extrinsic, domain-dependent tasks, consisting of sequence tagging and classification, show that the model is best in specific application scenarios, in contrast to the others. Furthermore, we show that our approach against catastrophic forgetting works, as the model is able to retrieve the previously trained domain-independent knowledge. The used dataset and trained model are made publicly available

cs.CR↗

arXiv2022

Reducing Information Overload: Because Even Security Experts Need to Blink

Computer Emergency Response Teams (CERTs) face increasing challenges processing the growing volume of security-related information. Daily manual analysis of threat reports, security advisories, and vulnerability announcements leads to information overload, contributing to burnout and attrition among security professionals. This work evaluates 196 combinations of clustering algorithms and embedding models across five security-related datasets to identify optimal approaches for automated information consolidation. We demonstrate that clustering can reduce information processing requirements by over 90% while maintaining semantic coherence, with deep clustering achieving homogeneity of 0.88 for security bug report (SBR) and partition-based clustering reaching 0.51 for advisory data. Our solution requires minimal configuration, preserves all data points, and processes new information within five minutes on consumer hardware. The findings suggest that clustering approaches can significantly enhance CERT operational efficiency, potentially saving over 3.750 work hours annually per analyst while maintaining analytical integrity. However, complex threat reports require careful parameter tuning to achieve acceptable performance, indicating areas for future optimization. The code is made available at https://github.com/PEASEC/reducing-information-overload.

cs.CR↗

arXiv2022

Common Vulnerability Scoring System Prediction based on Open Source Intelligence Information Sources

The number of newly published vulnerabilities is constantly increasing. Until now, the information available when a new vulnerability is published is manually assessed by experts using a Common Vulnerability Scoring System (CVSS) vector and score. This assessment is time consuming and requires expertise. Various works already try to predict CVSS vectors or scores using machine learning based on the textual descriptions of the vulnerability to enable faster assessment. However, for this purpose, previous works only use the texts available in databases such as National Vulnerability Database. With this work, the publicly available web pages referenced in the National Vulnerability Database are analyzed and made available as sources of texts through web scraping. A Deep Learning based method for predicting the CVSS vector is implemented and evaluated. The present work provides a classification of the National Vulnerability Database's reference texts based on the suitability and crawlability of their texts. While we identified the overall influence of the additional texts is negligible, we outperformed the state-of-the-art with our Deep Learning prediction models.

cs.CR↗

arXiv2022

Multi-Level Fine-Tuning, Data Augmentation, and Few-Shot Learning for Specialized Cyber Threat Intelligence

Gathering cyber threat intelligence from open sources is becoming increasingly important for maintaining and achieving a high level of security as systems become larger and more complex. However, these open sources are often subject to information overload. It is therefore useful to apply machine learning models that condense the amount of information to what is necessary. Yet, previous studies and applications have shown that existing classifiers are not able to extract specific information about emerging cybersecurity events due to their low generalization ability. Therefore, we propose a system to overcome this problem by training a new classifier for each new incident. Since this requires a lot of labelled data using standard training methods, we combine three different low-data regime techniques - transfer learning, data augmentation, and few-shot learning - to train a high-quality classifier from very few labelled instances. We evaluated our approach using a novel dataset derived from the Microsoft Exchange Server data breach of 2021 which was labelled by three experts. Our findings reveal an increase in F1 score of more than 21 points compared to standard training methods and more than 18 points compared to a state-of-the-art method in few-shot learning. Furthermore, the classifier trained with this method and 32 instances is only less than 5 F1 score points worse than a classifier trained with 1800 instances.

cs.CR↗

arXiv2021

A Survey on Data Augmentation for Text Classification

Data augmentation, the artificial creation of training data for machine learning by transformations, is a widely studied research field across machine learning disciplines. While it is useful for increasing a model's generalization capabilities, it can also address many other challenges and problems, from overcoming a limited amount of training data, to regularizing the objective, to limiting the amount data used to protect privacy. Based on a precise description of the goals and applications of data augmentation and a taxonomy for existing works, this survey is concerned with data augmentation methods for textual classification and aims to provide a concise and comprehensive overview for researchers and practitioners. Derived from the taxonomy, we divide more than 100 methods into 12 different groupings and give state-of-the-art references expounding which methods are highly promising by relating them to each other. Finally, research perspectives that may constitute a building block for future work are provided.

cs.CL↗

arXiv2021

Data Augmentation in Natural Language Processing: A Novel Text Generation Approach for Long and Short Text Classifiers

In many cases of machine learning, research suggests that the development of training data might have a higher relevance than the choice and modelling of classifiers themselves. Thus, data augmentation methods have been developed to improve classifiers by artificially created training data. In NLP, there is the challenge of establishing universal rules for text transformations which provide new linguistic patterns. In this paper, we present and evaluate a text generation method suitable to increase the performance of classifiers for long and short texts. We achieved promising improvements when evaluating short as well as long text tasks with the enhancement by our text generation method. Especially with regard to small data analytics, additive accuracy gains of up to 15.53% and 3.56% are achieved within a constructed low data regime, compared to the no augmentation baseline and another data augmentation technique. As the current track of these constructed regimes is not universally applicable, we also show major improvements in several real world low data tasks (up to +4.84 F1-score). Since we are evaluating the method from many perspectives (in total 11 datasets), we also observe situations where the method might not be suitable. We discuss implications and patterns for the successful application of our approach on different types of datasets.

cs.CL↗

arXiv2020

Empirical Insights for Designing Information and Communication Technology for International Disaster Response

Due to the increase in natural disasters in the past years, Disaster Response Organizations (DROs) are faced with the challenge of coping with more and larger operations. Currently appointed Information and Communications Technology (ICT) used for coordination and communication is sometimes outdated and does not scale, while novel technologies have the potential to greatly improve disaster response efficiency. To allow adoption of these novel technologies, ICT system designers have to take into account the particular needs of DROs and characteristics of International Disaster Response (IDR). This work attempts to bring the humanitarian and ICT communities closer together. In this work, we analyze IDR-related documents and conduct expert interviews. Using open coding, we extract empirical insights and translate the peculiarities of DRO coordination and operation into tangible ICT design requirements. This information is based on interviews with active IDR staff as well as DRO guidelines and reports. Ultimately, the goal of this paper is to serve as a reference for future ICT research endeavors to support and increase the efficiency of IDR operations.

cs.CY↗

arXiv2019

Cross-Media Usage of Social Big Data for Emergency Services and Volunteer Communities: Approaches, Development and Challenges of Multi-Platform Social Media Services

The use of social media is ubiquitous and nowadays well-established in our everyday life, but increasingly also before, during or after emergencies. The produced data is spread across several types of social media and can be used by different actors, such as emergency services or volunteer communities. There are already systems available that support the process of gathering, analysing and distributing information through social media. However, dependent on the goal of analysis, the analysis methods and available systems are limited based on technical or business-oriented restrictions. This paper presents the design of a cross-platform Social Media API, which was integrated and evaluated within multiple emergency scenarios. Based on the lessons learned, we outline the core challenges from the practical development and theoretical findings, focusing (1) cross-platform gathering and data management, (2) trustability and information quality, (3) tailorability and adjustable data operations, and (4) queries, performance, and technical development.

cs.SI↗