Morphological analysis

Abstract

In the previous chapters, we have seen that a lot of information about the potential tags of tokens in a text is found by lexicon lookup. Another, often complementary source of information is morphological analysis, i.e. the process of decomposing words into their constituents. The information about the individual constituents can be used to determine the necessary information about the word as a whole. Such information may range from basic wordclass information assigned from a fixed inventory of tags to structural information consisting of the relationships between components of the word further annotated with various features and their values (cf. Chapter 10). The English word “redness” could thus either be analysed as having the tag NN (singular noun) hiding its internal details, or be analysed by a suitable word grammar to have the structure Adj (red) + N (+ness) where the internal structure of the word has been made explicit.