When we started with the first example, and then explored some of the possibilities in the configurations for text analysis of a field type, we saw that it is important to handle terms correctly in Lucene and Solr. We saw this again when using facets as we could obtain very different results by simply handling terms on the same data with different approaches. And we have often seen that terms are very important for the construction of the final score of the results. We don't want to go too far in handling terms directly. Instead, we prefer to use higher-level approach, but this approach can still give some idea in case you want to study more on this later.