Our system for annotation of articles is named “Text Detective”

Text Detective is then able to tag every word in the sentence according to biological relevant categories. For instance, chemical compounds are recognized and labelled. The identification of “central words” (also known as “core terms”) is a key step in this process (words such as “receptor”, “kinase”, “transporter”, etc). For this purpose, we have built a lexicon and used some carefully curated rules. Also “types” are tagged (words such as “alpha”, “a1”, “c”, “12”, “TNF”), since they may define the exact identity of the gene (distinguishing between “interferon alpha” and “interferon gamma”, for instance). These “type” words are recognized after a set of carefully designed rules (presence of capital letters, numbers, Greek letters, etc.). Notice that gene symbols (such as “TNFalpha”) are also tagged as “type”.