search

dogatuncay / Twitter-Emoticons

Find out how using the emoticons as classifiers affects the classification algorithms’ accuracy
1 stars 0 forks source link

Could you provide me some data in this format? #1

Open lcmuch opened 9 years ago

lcmuch commented 9 years ago

Here is a simple Naive Bayes test data

positive or negative, whether it contains word a1, wether it contains word a2, ....

0,1 0 0 0,1 0 0 1,0 1 0 1,0 1 0

So we need to have

  1. Find a list of important words for sentiment analysis first;
  2. Then find tweets with :) and set them to be positive tweets Find tweets with :( and set them to be negative tweets
  3. Search target important words and same the frequency to list
  4. Run my app to know the accuracy by Naive Bayes classification
lcmuch commented 9 years ago

By the way, I do not know how to add commends to MLProject...