Vera Schmitt

Vera Schmitt
Technische Universität Berlin | TUB · Department of Software Engineering and Theoretical Computer Science

Dr. rer. nat.
Head of XplaiNLP research group (TU Berlin) https://www.tu.berlin/en/qu/research/research-groups/xplainlp

About

31
Publications
2,792
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
82
Citations
Introduction
NLP - developing and fine-tuning LLMs for mis-and disinformation, and hate speech detection HCI - developing intelligent decision-support systems for enhanced decision-making, developing xAI features for transparent AI-recommendations Usable Privacy/Privacy by Design Legal and ethical implications of regulations on AI Ethical Implications of Fake News/Hate Speech Detection
Additional affiliations
May 2023 - present
Deutsches Forschungszentrum für Künstliche Intelligenz
Position
  • Guest Researcher

Publications

Publications (31)
Preprint
Full-text available
Counterfactual examples are widely used in natural language processing (NLP) as valuable data to improve models, and in explainable artificial intelligence (XAI) to understand model behavior. The automated generation of counterfactual examples remains a challenging task even for large language models (LLMs), despite their impressive performance on...
Preprint
Full-text available
In this work, we introduce a methodology for alignment designed to enhance the ability of large language models (LLMs) to articulate their reasoning (self-explanation) even in the absence of annotated rationale explanations. Our alignment methodology comprises three key components: explanation quality assessment, self-instruction dataset generation...
Preprint
Full-text available
Natural language explanations (NLEs) are vital for elucidating the reasoning behind large language model (LLM) decisions. Many techniques have been developed to generate NLEs using LLMs. However, like humans, LLMs might not always produce optimal NLEs on first attempt. Inspired by human learning processes, we introduce Cross-Refine, which employs r...
Conference Paper
Full-text available
With the rise of Large Generative AI Models (LGAIMs), disinformation online has become more concerning than ever before. Within the super-election year 2024, the influence of mis- and disinformation can severely influence public opinion. To combat the increasing amount of disinformation online, humans need to be supported by AI-based tools to incre...
Conference Paper
Full-text available
Numerous smartphone and web applications rely on personal information to analyze user behavior, primarily for targeted advertising and the enhancement of personalized features. However, these applications often provide users with limited choices: either accept their privacy policies or refrain from using the services altogether. Consequently, the p...
Chapter
Full-text available
Numerous smartphone and web applications rely on personal information to analyze user behavior, primarily for targeted advertising and the enhancement of personalized features. However, these applications often provide users with limited choices: either accept their privacy policies or refrain from using the services altogether. Consequently, the p...
Chapter
To achieve the 17 Sustainable Development Goals (SDGs), it is essential to monitor and measure the contributions of various stakeholders, including companies. However, existing methods for automatically analyzing SDG contributions in sustainability reports (SRs) focus on sentence-based classification, discerning SDG context spread over several sent...
Article
Full-text available
Nowadays, smartphones are equipped with various sensors collecting a huge amount of sensitive personal information about their users. However, for smartphone users, it remains hidden, and sensitive information is accessed by used applications and data requesters. Moreover, governmental institutions have no means to verify if applications requesting...
Chapter
Full-text available
Various smartphone and web applications use personal information to estimate the user’s behaviour among others for targeted advertising and improvement of personalized applications. Often applications and web services offer only two choices, either accept their privacy policies or not use the services. Hereby, the general scenario is to pay applica...
Chapter
Nowadays, smartphones are equipped with various sensors collecting a huge amount of sensitive personal information about their users. However, for smartphone users, it remains hidden, and sensitive information is accessed by used applications and data requestors. Moreover, governmental institutions have no means to verify if applications requesting...
Conference Paper
Full-text available
In 2015, The United Nations (UN) provided a blueprint for sustainable development in various domains. This Blueprint described 17 Sustainable Development Goals (SDGs), such as "No Poverty", "Zero Hunger", or "Gender Equality". Subsequently, many companies have started publishing yearly sustainability reports, explaining their efforts with respect t...
Conference Paper
Full-text available
Automated text readability assessment is the process of assigning a number to the level of difficulty of a piece of text automatically. Machine learning and natural language processing techniques made it possible to measure the readability and complexity of the fast-growing textual content on the web. In this paper, we proposed a multi-task learnin...
Conference Paper
Full-text available
Privacy can be approached from different perspectives. Within this research, the cognate-based definition of privacy is examined with respect to the influence of elemental traits of individuals, such as hungriness. Hungriness is examined in relation to privacy constructs , namely, privacy concern, privacy literacy, and trust and risk beliefs. This...
Chapter
In our daily life, we make extensive use of location-based services when searching for a restaurant nearby, searching for an address we want to visit, or searching for the best route to drive. Location information is highly sensitive personal information that users share without the awareness of being continuously tracked by various apps on their s...
Conference Paper
Android applications request specific permissions from users during the installations to perform required functionalities by accessing system resources and personal information. Usually, users must approve the permissions requested by applications (apps) during the installation process and before the apps can collect privacy- or security-relevant i...
Preprint
Full-text available
The detection of hate speech online has become an important task, as offensive language such as hurtful, obscene and insulting content can harm marginalized people or groups. This paper presents TU Berlin team experiments and results on the task 1A and 1B of the shared task on hate speech and offensive content identification in Indo-European langua...
Conference Paper
Full-text available
Companies frequently make claims about positive impacts on the environment, but these claims are not always valid. In this sense, “greenwashing” is used when a company states a false claim about its products and practices being environmentally friendly. It is an important issue and has attracted the attention of policymakers to strengthen consumer...
Conference Paper
Full-text available
In April 2021 the European Commission (EC) proposed a new regulation to establish a regulatory structure for the risk assessment of Artificial Intelligence (AI) systems and applications. The intended goal of initiating a harmonised legal framework for the European Union (EU) poses new challenges in developing countermeasures for hate speech and fak...
Conference Paper
Full-text available
The standard approach of Notice and Choice does not provide sufficient control over personal privacy preferences. A more granular analysis of privacy preferences is needed where the monetary valuation of different data types can contribute to the understanding of individual privacy concerns of personal information. The question of how much consumer...

Network

Cited By