Creating a candidate set of key phrases

• In keyphrase extraction, the question is how to make candidates

• RAKE chops the stop word as a delimiter.

• TextRank first filters out all but nouns and adjectives

  • Phrase-based TF-IDF uses only the longest noun phrase

• All substrings are candidates. - Development of a system for extracting keywords in unexplored text information

• I want to include strings enclosed in parentheses as keyword candidates.

• When trying to find key phrases with an approach that calculates scores for the candidates and takes the big ones

  • EmbedRank uses similarity to the original document, so the text itself must not be included in the candidates.

---

This page is auto-translated from /nishio/キーフレーズ候補集合の作成 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.