For example, if “lecture” is extracted as a keyword and then the statement “make lecture materials” is made, if keyword extraction extracts the range of “lecture materials” as a keyword, this would result in “B” being extracted independently of “A,” which is not appropriate. In this case, “lecture” should also be regarded as appearing again.

Added ✅Recurrences of short keywords are not treated as keywords.

make_response.py

# Consider substrings that are not selected in the keyword extraction but appear as keyword occurrences.
for k in env.scored_keyphrases:
    if k not in keyphrase and k in s:
        keyphrase[k] = min(100, env.scored_keyphrases[k])

This page is auto-translated from /nishio/✅出現してるがキーワード抽出で選ばれてない部分文字列をキーワード出現とみなす using DeepL. If you looks something interesting but the auto-translated English is not good enough to understand it, feel free to let me know at @nishio_en. I’m very happy to spread my thought to non-Japanese readers.