Subjective Content Descriptions: Semantic Representation Learning

Investigator: Prof. Dr. Ralf Möller

Research Associate: Felix Kuhr

An agent in pursuit of a task may work with an individual collection of documents, which is known as a corpus. Each document is associated with additional data providing location-specific subjective content descriptions. Manually creating subjective content descriptions for documents is a time-consuming task. Therefore, this project investigates a domain-adaptation approach for context- and location-specific enrichment of documents from one corpus using already existing subjective content descriptions associated to documents of another corpus. A case study shows the effectiveness of our approach using a source corpus containing documents which are associated with subjective content descriptions and a target corpus containing only documents.