search

cyhex / streamcrab

Real-Time, Twitter sentiment analyzer engine
http:/www.streamcrab.com
144 stars 49 forks source link

Not sure what changes to make to twitterCollector.py #3

Closed dwmcqueen closed 12 years ago

dwmcqueen commented 12 years ago

What fields should be replaced? I see username / password - is this Twitter username / password? Anything else?

cyhex commented 12 years ago

i have made some changes to twitterCollector.py, take a look.

but the only thing that you have to change is the username/password for your twitter account.

dwmcqueen commented 12 years ago

How does it avoid using Oauth authentication?

cyhex commented 12 years ago

the stream api supports basic http auth....

dwmcqueen commented 12 years ago

Perfect. Looked at the code and it makes sense. Any tips on what to put in here:

""" List of keywords to track"""
trackingK = [':)']

I assume keywords that lead to positive tweets?

cyhex commented 12 years ago

yes, you can also use smiles as keywords. please remember that you have to have about the same size positive and negative tweets db.

dwmcqueen reply@reply.github.com wrote:

Perfect. Looked at the code and it makes sense. Any tips on what to put in here:

""" List of keywords to track""" trackingK = [':)']

I assume keywords that lead to positive tweets?

Reply to this email directly or view it on GitHub: https://github.com/cyhex/smm/issues/3#issuecomment-2455283

dwmcqueen commented 12 years ago

Thank you. Comma delimited (sorry for all questions)?

On Wed, Oct 19, 2011 at 9:33 AM, cyhex < reply@reply.github.com>wrote:

yes, you can also use smiles as keywords. please remember that you have to have about the same size positive and negative tweets db.

dwmcqueen reply@reply.github.com wrote:

Perfect. Looked at the code and it makes sense. Any tips on what to put in here:

""" List of keywords to track""" trackingK = [':)']

I assume keywords that lead to positive tweets?

Reply to this email directly or view it on GitHub: https://github.com/cyhex/smm/issues/3#issuecomment-2455283

Reply to this email directly or view it on GitHub: https://github.com/cyhex/smm/issues/3#issuecomment-2456451

Duncan McQueen OpenThinking Systems, LLC http://www.openthinkingsystems.com

cyhex commented 12 years ago

you need to collect 2 different data sets (files) one for positive one for negative, take a look at collector / trainer / tweetClassifier.py that will give you a better understanding.