#SchizoidMan

• 

  our process

• 

  output of calling schizoidman module

Inspiration

The idea of artificial agents being aloof and lacking understanding of what it means to be sentient (have emotion), is not new. One may thus describe these unfeeling AI bots as 'schizoid' ('anti-social'). We set out to change this. Before effective communication (the AI bots to responding to the complex human) there needs to be understanding of what is truly meant by the human language and acknowledgement of the context of the human psyche (personality). This project thus aims to improve AI chatbot capabilities in respect of the above mentioned concepts. Application in the business context will have a multitude of benefits in terms of enhancing customer relationships and service.

What it does

After downloading schizoidman.py, by creating a notebook within the .py file, the user can type : import schizoidman to see a demonstration. Twitter freshdesk tickets add info is used to trace the customers profile on twitter, their tweets are then analyzed to determine their personality.

How we built it

Connected Freshdesk to python, twitter to python, did machine learning and tried to build and deploy a streamlit app on AWS.

Challenges we ran into

Deploying the machine learning app.

Accomplishments that we're proud of

The video we created.

What we learned

Keep code simple.

What's next for #SchizoidMan

Customer segmentation, triggered by freshdesk ticket content, gamification to promote movement beween less favorable and ideal customer segments.

Built With

• freshdesk.api
• gethandles
• nltk.download('vader-lexicon')
• nltk.sentiment.vader
• pandas
• pycaret.classification
• sentimentintensityanalyzer
• snscrape.modules.twitter
• sntwitter
• twitterusernameviauserid