We propose a novel method to detect the corresponding long forms (LFs) of short forms (SFs) from biomedical text. The proposed method is differentiated from others as follows: it incorporates lexical analysis techniques into supervised learning for extracting abbreviations; it utilises text-chunking...