Semantic analysis


Beyond the limitation of textual analysis based on the frequency of recurrent words and phrases, and the limitation of tagging analysis based on associating some occasional words that may convey some meaning to a reader, but is prone for instance to misspelling and subjective inconsistency, Semantic Analysis (SA) approaches the knowledge structuring and search issues from a data processing point of view by means of semantic “tokens” that provide context to the natural language. The traditional keyword approach is replaced by a concept driven one, where concepts can be identified through different semantic analysis approaches: from statistical ones, to information retrieval, domain knowledge driven, and predicate logic approaches.

A number of research activities are ongoing at ISIN in the area of semantic analysis applied to various domains. These activities includes research student projects, aiming at developing and strengthen our expertise in the field. The main focused topics considered so far include but are not limited to:

  • Natural Language Processing (NLP)
  • Pshycological profiling
  • User profiling
  • Statistical approaches to data analysis
  • Semantic web and Linked Open Data

Various international and national projects in this areas are currently carried out at ISIN providing a growing experience and competence in related topics.

Further Information

Lorenzo Sommaruga, Professor

st.wwwsupsi@supsi.ch