A5002799905- Semantic Web and Ontologies
- Advanced Text Analysis Techniques
- Natural Language Processing Techniques
- Scientific Computing and Data Management
- Biomedical Text Mining and Ontologies
- Topic Modeling
- Data Quality and Management
- Research Data Management Practices
- Library Science and Information Systems
- Advanced Database Systems and Queries
- Libraries and Information Services
- Service-Oriented Architecture and Web Services
- Distributed and Parallel Computing Systems
- Data Visualization and Analytics
- Web Data Mining and Analysis
- Digital Humanities and Scholarship
- Data Mining Algorithms and Applications
- Auction Theory and Applications
- Bullying, Victimization, and Aggression
- Video Analysis and Summarization
- Advanced Malware Detection Techniques
- Hate Speech and Cyberbullying Detection
- Mathematics, Computing, and Information Processing
- Advanced Graph Neural Networks
- scientometrics and bibliometrics research
Mannheim University of Applied Sciences
2023-2024
Stuttgart Media University
2015-2022
University of Stuttgart
2016-2017
University of Mannheim
2007-2015
Cyberbullying is a disturbing online misbehaviour with troubling consequences. It appears in different forms, and most of the social networks, it textual format. Automatic detection such incidents requires intelligent systems. Most existing studies have approached this problem conventional machine learning models majority developed these are adaptable to single network at time. In recent studies, deep based found their way cyberbullying incidents, claiming that they can overcome limitations...
Argumentation is arguably one of the central features scientific language. We present ArguminSci, an easy-to-use tool that analyzes argumentation and other rhetorical aspects writing, which we collectively dub scitorics. The main aspect focus on fine-grained argumentative analysis text through identification argument components. functionality ArguminSci accessible via three interfaces: as a command line tool, RESTful application programming interface, web application.
The "wisdom of crowds" is accomplishing tasks that are cumbersome for individuals yet cannot be fully automated by means specialized computer algorithms. One such task the construction thesauri and other types concept hierarchies. Human expert feedback on relatedness relative generality terms, however, can aggregated to dynamically construct evolving InPhO (Indiana Philosophy Ontology) project bootstraps from volunteer users unskilled in ontology design into a precise representation specific...
Citation graphs and indices underpin most bibliometric analyses. However, measures derived from citation do not provide insights into qualitative aspects of scientific publications. In this work, we aim to semantically characterize citations in terms polarity purpose. We frame purpose detection as classification tasks investigate the performance convolutional networks with general domain-specific word embeddings on these tasks. Our best performing model outperforms previously reported...
Exponential growth in the number of scientific publications yields need for effective automatic analysis rhetorical aspects writing. Acknowledging argumentative nature text, this work we investigate link between structure and such as discourse categories or citation contexts. To end, (1) augment a corpus annotated with four layers rhetoric annotations argumentation (2) neural multi-task learning architectures combining argument extraction set classification tasks. By coupling classifiers...
Abstract The number of scientific publications nowadays is rapidly increasing, causing information overload for researchers and making it hard scholars to keep up date with current trends lines work. Recent work has tried address this problem by developing methods automated summarization in the scholarly domain, but concentrated so far only on monolingual settings, primarily English. In paper, we consequently explore how state-of-the-art neural abstract models based a multilingual...
Citations play a crucial role in the scientific discourse, information retrieval, and bibliometrics. Many initiatives are currently promoting idea of having free open citation data. Creation data, however, is not part cataloging workflow libraries nowadays.
The DM2E dataset is a five-star providing metadata and links for direct access to digitized content from various cultural heritage institutions across Europe. data model true specialization of the Europeana Data Model reflects specific requirements domain manuscript s old prints, as well developers who want create applications on top data. One such application scholarly research platform Digital Humanities that was created part project can be seen reference implementation. Linked API...
Abstract Many disciplines, including the broad Field of Information (iField), offer Data Science (DS) programs. There have been significant efforts exploring an individual discipline's identity and unique contributions to broader DS education landscape. To advance in iField, iSchool Curriculum Committee (iDSCC) was formed charged with building recommending a framework for iSchools. This paper reports on research process findings series studies address important questions: What is iField...
Purpose The purpose of this work is to explore the new possibilities enabled by recent introduction RDF-star, an extension that allows for statements about within Resource Description Framework (RDF). Alongside Named Graphs, approach offers opportunities leverage a meta-level data modeling and applications. Design/methodology/approach In extended paper, authors build onto three use cases published in previous paper: (1) provide provenance information, (2) maintain backwards compatibility...
For data practitioners embracing the world of RDF and Linked Data, openness flexibility is a mixed blessing. them, validation according to predefined constraints much sought-after feature, particularly as this taken for granted in XML world. Based on our work DCMI Application Profiles Task Group cooperation with W3C Data Shapes Working Group, we published by today 81 types that are required various stakeholders applications. These constraint form basis investigate role reasoning different...
The use of thesaurus-based indexing is a common approach for increasing the performance document retrieval. With growing amount documents available, manual not feasible option. Statistical methods automated are an attractive alternative. We argue that quality thesaurus used as basis in regard to its ability adequately cover contents be indexed crucial importance inautomatic because there no human loop can spot and avoid errors. propose method evaluation based on combination statistical...