Understanding KJ Method in the Context of Machine Learning

What happens when we consider the KJ Method in the context of Machine Learning?

• (Task: Class classification)

  • (I don’t have much knowledge about both, so I can’t say anything definite)

  • Connect it to Unsupervised Learning and related texts.

  • Without the context in the first place, it cannot be vectorized.

    • I wonder how this task is done in Natural Language Processing.

  • Is the KJ Method originally an intellectual production method for humans?

  • “Feature extraction” may correspond to having common context/background information in Language.

    • It becomes easier to interpret.

  • When it comes to finding features with large dispersion, is that the KJ Method?

    • Talking about the same topic.