It is well known that parsing accuracy suffers when a model is applied to out-of-domain data. It is also known that the most beneficial data to parse a given domain is data that ...
The goal of sentiment prediction is to automatically identify whether a given piece of text expresses positive or negative opinion towards a topic of interest. One can pose sentim...
The SLIF project combines text-mining and image processing to extract structured information from biomedical literature. SLIF extracts images and their captions from published pap...
Automatic classification of documents is an important area of research with many applications in the fields of document searching, forensics and others. Methods to perform class...
Background: When term ambiguity and variability are very high, dictionary-based Named Entity Recognition (NER) is not an ideal solution even though large-scale terminological reso...
Yutaka Sasaki, Yoshimasa Tsuruoka, John McNaught, ...