FUSION
FUnctionality Sharing In Open eNvironments
Heinz Nixdorf Chair for Distributed Information Systems

Disambiguation of Ontological Concepts for Semantic Datatset Annotation (Bachelor or Master Thesis/ German or English)

 

Semantic Annotations are additional information on textual or structrued data that link terms to ontological concepts. This helps to incorporate additional information into the textual context. This text mining technique is getting increasing attention since it is required in text mining applications, semantic search or question answering systems. A main issue in creating semantic annotations is disambiguation, which means to define a clear context for a term or phrase. For instance, a Jaguar could be a car or an animal. Depending on the surrounding terms, a system could try to disambiguate the context and could select the correct ontology or vocabulary. The main task of the thesis is to develop a ranking function for annotation concept candidates based on the given context. The outcome should be a ranked list of suitable ontological concepts for a given term or phrase. The proposed solution is supposed to be evaluated against a manually annotated corpus of datasets. Datasets and ontologies will be provided by the supervisor. The thesis can be written in German or English.

What are the tasks?

  • State of the art of disambiguation approaches
  • Development of a ranking function for ontological concepts
  • Evaluation of the developed solution

What do we expect?

  • Programming knowledge in JAVA
  • Foundations of semantic web technologies

What do we offer?

  • Training support in text mining foundations including tools such as GATE (https://gate.ac.uk/)
  • A nice working environment on the 17th floor in the Jentower, Jena with a marvelous view over the city
  • An interdisciplinary and international team
  • The opportunity for a continued employment as student assistant during master studies

Start: April 2019 or later| Supervisor: felicitas.loeffler@uni-jena.de