Skip to main content
Figure 2 | Journal of Cheminformatics

Figure 2

From: OSCAR4: a flexible architecture for chemical text-mining

Figure 2

OSCAR3 markup displaying recognised chemical entities (CM). A mouse-over action on an annotated term displays the associated metadata, in this case for 2,5-dichlorobenzylamine, and displays an image representing the structure generated by the Chemistry Development Kit (CDK) [3537] (right). OSCAR3 concentrated on the identification and interpretation of chemical entities in text (named entity recognition, NER). The primary purpose was to identify and extract the following types of object: chemicals (CM), ontology terms (ONT; looked-up from ChEBI [3840], FIX [41] and REX [42]etc.), reactions (RN; as identified by linguistic constructs, e.g. "methylated"), chemical adjectives (CJ) mainly formed from chemical nouns), enzymes (ASE) and chemical prefixes (CPR), highlighted in different colours. These concepts are maintained in OSCAR4.

Back to article page