Inspiration

Cognitive behavioral therapy (CBT) is a problem-focused psychotherapy treatment that teaches clients to identify and change unhelpful thinking patterns and behaviors. CBT has been successful in treating anxiety, depression, eating disorders, substance abuse disorders, marital problems, psychosis and personality disorders (Beck Institute for Cognitive Behavior Therapy, 2019).

CBT uses many different tools to help people change their thoughts and behavior. These include:

Changing thinking patterns. People are taught to recognize negative thoughts and replace them with more positive ones. This helps them think more realistically about situations.

Monitoring daily activities. Clients keep track of all their daily activities and how they feel about them by using a diary or journal. They can then use this information to determine the situations that lead to their negative thoughts and feelings.

Problem-solving skills training. Clients are given skills training so they can solve problems on their own.

Social skills training. Clients learn how to improve their communication skills and make friends more easily by practicing social interactions with each other in group therapy sessions (Beck Institute for Cognitive Behavior Therapy, 2019).

What it does

The rise of A. I can be seen in many fields and it had shown great results. so why not merge A. I with Cognitive Behavioral Therapy (CBT) and increase the access to mental healthcare especially In a country like India where we have very less regard for mental health also have very less mental health doctors.

The World Health Organization, in 2019, had estimated that 7.5% of Indians were affected by mental health disorders. This number would likely go up significantly because of the pandemic. In a country with a deep mental health crisis, and with only above two mental health beds for every 1,00,000 population, mental health social workers, therapists and counsellors become a significant bridge in community health.

Using this kind of bot we can deploy these in mobile phones or websites and may be able to provide mental healthcare access in remote areas or to anybody who is ashamed of society to go to the doctor. Increasing awareness about mental health would be great and AISHIKA can help by making access more flexible.

How we built it

  1. I built up an intents JSON file that defines certain intentions that could occur during the interactions with our chatbot. Within this intents JSON file, alongside each intents tag and pattern, there will be responses. Libraries we are going to use are - json, string, random, nltk, numpy, wordnetlemmatizer, tensorflow.
  2. First we going to use Lemmatizer method - It helps understand the contextual meaning. Applying Bag of words - The bag-of-words model is commonly used in methods of document classification where the (frequency of) occurrence of each word is used as a feature for training a classifier Sequential Neural Net - A Sequential model is appropriate for a plain stack of layers where each layer has exactly one input tensor and one output tensor. We are going to use to 3 dense layer and two 2 dropout to reduce overfitting.

Challenges we ran into

One of the great challenges is to merge a chatbot with Cognitive Behavioral Therapy. So we need to learn a lot about this because what questions to ask is a great part of a CBT and especially which one to ask and then analyse the answer and return a well-matched response.

From this learning, we understood that key factors of a CBT session included:

“targeting emotions by changing thoughts and behaviours that are contributing to distressing emotions” Creating a strong therapeutic relationship through empathy (validating the patient’s experience), genuineness (being authentic), - positive regard (respect) Collaborative work of skill acquisition and homework to mould positive thought patterns Teaching skills rather than just talking Most importantly, the therapist must have collaborative, assertive, and nonjudgemental.

Accomplishments that we're proud of

What's next for Aishika

Integration with a voice Assistant

Share this project:

Updates