Semantic Normalization
In the context of SolrSherlock, a way to think about Semantic Normalization is to think in terms of term substitution. A trivial example would be to take find all of the various ways that carbon dioxide is named in text being harvested and substitute one and only one name, say co2.
The context in which this process is performed is that of identifying topics in the topic map which are unambiguously named with that term. This then allows to represent free text in terms of the topics entailed by the words.