So i wrote a script that gets sentiment from any tweets containing a key word/words. I searched somalis and surprise surprise.
Last edited:
Proof that somalis have a negative brand
- Thread starter Siddhartha
- Start date
- Status
Who is denying Somalis have a negative brand? You just wasted your valuable labour for nothing. 
Who is denying Somalis have a negative brand? You just wasted your valuable labour for nothing.
Lol this is just a side effect created it for other reasons.
Twitter isn't the best place to get a real feel on how the world thinks about us or any other group.
Slightly of topic.
Here's a cool with 101 I found of sentiment analysis: https://monkeylearn.com/sentiment-analysis/
Slightly of topic.
Here's a cool with 101 I found of sentiment analysis: https://monkeylearn.com/sentiment-analysis/
Cool link, in this case i used rule based approach Vader from NLTK. nlp with AI is fascinating, its amazing how words can be represented in 1s and 0s
Wasn't aware of that library. Seems pretty useful. I guess rule based systems could be effective within a limited role.
Always assumed that to get any effective sentiment analysis going you needed to know all this advanced math concepts like group/information theory and so on.
I guess the above is only needed for deep learning models where you want to pull out the nuances in tweets...i.e. getting into messy semantics.
The subject is still new so who knows where it'll go.
All I know is that only large marketing or consulting firms need people knowledgeable about sentiment tracking atm.
Nope i used Python. JavaScript is not a general purpose programming language and nodejs is a joke lol.
Why are you a self haters
- Status
Trending
-
Somali man getting cooked on twitter for not liking his wife’s affection- Started by empressjamila
- Replies: 52
-
-
