Limitations
The querying capabilities of this Python package is highly dependent on the navigational structure of each source in the query pool. If a source modifies its navigational structure then extraction from that specific source will likely have some challenges. The maintainers of WordHoard will correct these navigational extraction issues when they are discovered in periodic testing or reported as an issue.
Another package limitation is the accuracy of certain antonyms or synonyms provided by a source. For instance WordHippo (a source still under evaluation) provides an overabundance of synonyms for a given word. In most cases the words provided have almost no relationship to the word inputted.
For example:
from wordhoard import Synonyms
synonym = Synonyms('banana')
results = synonym.query_wordhippo() # debugging source
print(results)
['actor', 'buffoon', 'card', 'clown', 'comedian', 'comic', 'cutup', 'droll', 'farceur', 'fool', 'funster',
'gagster', 'humorist', 'jester', 'jokesmith', 'jokester', 'jook-sing', 'josher', 'kidder', 'life of the party',
'prankster', 'punster', 'quipster', 'second banana', 'stand-up comic', 'stooge', 'straight person', 'top banana',
'trickster', 'twinkie', 'wag', 'wisecracker', 'wit']
It is hard for an application like WordHoard to intuitively extract the words, which are actually synonyms from the WordHippo results. To counter this WordHippo was removed for the Synonyms module.
Additionally, some sources provide near antonyms or near synonyms for a specific word. The maintainers of WordHoard have made a best effort to remove these types of erroneous responses from the output of any given query.