The search finds tokens in all accepted documents of the PMB with the help of the several filters. Filtering is done based on the predefined list of values. An exception is a search according to a token/lemma, where a search term is interpreted as a MySQL regex and (case-insensitive) matches to the token or the lemma of a word token. Use explicit start (^) and end ($), e.g. '^read$', to search for exact words tokens and avoid the results including the term as a substring, e.g. 'already'. If you want to filter by semantic class, try the semantic lexicon. To search for multi-word expressions, such as New York, which are seen as a single token, use the tilde, e.g. 'New~York'. read more>>>

document category pos roles token/lemma sense semtag snippet subcorpus part