Clustering with Bert Embeddings

Поділитися
Вставка
  • Опубліковано 26 жов 2024

КОМЕНТАРІ • 12

  • @vasanthpragash854
    @vasanthpragash854 Рік тому +1

    I really like the second cluster. Its funny to see twitter at its worst.

  • @Jxxxxxxxxxxxxxxxxxxx
    @Jxxxxxxxxxxxxxxxxxxx Рік тому +2

    can this approach be useful to group google reviews into categories

  • @SofiaAmelSiyoucefSophie
    @SofiaAmelSiyoucefSophie Рік тому +2

    how do you measure the quality of this clustering ? thank you

    • @deepakjohnreji
      @deepakjohnreji  Рік тому +1

      Hi Sofia, this link might help, couple of statistical techniques are mentioned here towardsdatascience.com/clustering-evaluation-strategies-98a4006fcfc

  • @LowVi
    @LowVi 2 роки тому +1

    THANK YOU SO MUCH

  • @ting-chiehhuang6937
    @ting-chiehhuang6937 Рік тому

    can this be used in address grouping??

    • @deepakjohnreji
      @deepakjohnreji  Рік тому

      I don't think this approach will work for that.

  • @priyankaghadge6758
    @priyankaghadge6758 2 роки тому

    Can we use above code for sentence similarites

    • @deepakjohnreji
      @deepakjohnreji  2 роки тому

      Yes, the underlying concept can be used, But I would suggest you can refer to this code implementation for Sentence similarity
      Bert implementation: ua-cam.com/video/x5iYb0q2l2g/v-deo.html
      Elmo Implementation: ua-cam.com/video/Lvv4n5_Jjj0/v-deo.html

  • @elchino356
    @elchino356 2 роки тому +1

    Great, thanks a lot

  • @giridharnair01
    @giridharnair01 2 роки тому +1

    Good article bro