"naive bayes algorithm"

RamanRaman Member Posts: 8 Contributor I
edited June 2019 in Help
Hello Sir,
I am Research Scholar. I m pursuing my master degree and my thesis topic is Sentiment Analysis of reviews using naive bayes algorithm by Rapidminer. Can you provide me Naive Bayes algorithm???Β  I need it for my research.
My email id is [email protected]
Tagged:

Answers

  • mschmitzmschmitz Administrator, Moderator, Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert, University Professor Posts: 2,316  RM Data Scientist
    - Head of Data Science Services at RapidMiner -
    Dortmund, Germany
  • RamanRaman Member Posts: 8 Contributor I
    Sir,
    In my project, i have classified reviews by using naive bayes algorithm in Rapidminer and in my thesis, i have to improve that technique( improve or enhance naive bayes algorithm) to get better results and increase its accuracy.
    or can you provide me any other alternative ?
  • mschmitzmschmitz Administrator, Moderator, Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert, University Professor Posts: 2,316  RM Data Scientist
    Linear SVM with optimized C?
    - Head of Data Science Services at RapidMiner -
    Dortmund, Germany
  • RamanRaman Member Posts: 8 Contributor I
    Linear SVM classifier is already choosen by another research scholar in my college. Can you give me any idea how can i improve or change naive bayes algorithm to increase the accuracy ???
  • mschmitzmschmitz Administrator, Moderator, Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert, University Professor Posts: 2,316  RM Data Scientist
    oh, the trick is not the best algorithm. The trick is do to good feature engeneering.

    How did you to your preprocessing?
    - Head of Data Science Services at RapidMiner -
    Dortmund, Germany
  • RamanRaman Member Posts: 8 Contributor I
    first of all, i have done tokenisation, then filter stopwords, filter tokens, stem (Porter), and then tranform cases.
    image
  • RamanRaman Member Posts: 8 Contributor I
    hello sir...,
    what should i do next ?
  • mschmitzmschmitz Administrator, Moderator, Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert, University Professor Posts: 2,316  RM Data Scientist
    have you tried n-gram? Might be very useful for a naive bayes. This way you can incoperate correlation efeects.

    What if you try to optimize different options in your prep?
    - Head of Data Science Services at RapidMiner -
    Dortmund, Germany
    Thomas_Ott
Sign In or Register to comment.