Expected learning outcomes
- Understand the difference between vectors, matrices, vector space, vector field.
- Understand bicomponential theories of word meaning.
- Get a grasp of term weighting to distribute content in vector space.
- See the link between distributed content vs. lexical fields in semantics.
What is the target audience?
- Digital Preservation and Digital Humanities practitioners who are interested in investigating the deployment of semantics and relevant cutting-edge technologies.
- Researchers working on semantics who are interested in investigating the application of relevant methods and technologies in a Humanities field.
Level of advancement/ prerequisites
This recorded lecture is part of a larger module “Dynamics of knowledge organisation”: http://pericles-project.eu/training-module/dynamics-of-knowledge-organisation/
This is chapter 2 of the complete set:
- Semantics and DP: basic concepts, theories and trends
- Vectors and matrices: word meaning for advanced access
- Vector fields: a new approach to evolving semantics
- The Semantic Web & the Emergence of Ontologies
- Semantic Technologies & DP
- Evolving Semantics: Ontology Evolution & Semantic Drifts
Time required for completion25 min
Turney and Pantel’s paper is a very good overview of how word meaning can be captured by contextualized index term occurrence and co-occurrence frequencies. At the same time, being a review article, it is also technical and requires a level of technical understanding and the ability to generalize.
Turney, P.D., Pantel, P. 2010. From Frequency to Meaning: Vector Space Models of Semantics.Journal of Artificial Intelligence Research (37), 141-188.