Document keywords are associated to documents as summarized versions of the documents' content. Considering that the number of documents is quickly growing every day, the availability of these keywords is very important. Although, usually keywords are manually written. This motivated us to work on an approach to change this manual procedure for an automatic one. This paper presents a language independent approach that extracts the most relevant Multiword Expressions and single words from documents and propose them to describe the core content of each document.
|Title of host publication||International Conference on Computational Linguistics|
|Publication status||Published - 1 Jan 2010|
|Event||International Conference on Computational Linguistics (COLING) - |
Duration: 1 Jan 2010 → …
|Conference||International Conference on Computational Linguistics (COLING)|
|Period||1/01/10 → …|