Journal-Bot

• 

  Journal Bot UI

Inspiration

Around the world we are facing a mental health crisis, and there are alarming statistics regarding rates of anxiety and depression among youth and young adults.

In Canada, approximately 1.6 million children and teens are grappling with mental health disorders, with significant delays in treatment for tens of thousands.

Specifically, there is a disproportionate impact on Black, Indigenous, and other racialized individuals, as well as those questioning their gender and sexual identities, who often encounter additional barriers in accessing mental health care.

Journaling is proven to be beneficial for self-reflection and emotional well-being. However, when grappling with intense emotions, it can be challenging to navigate the process alone.

What it does

Our team focuses on innovating for inclusive healthcare.

Journal Bot offers a safe, confidential space for self-exploration through conversational journaling. By leveraging AI technology, Journal Bot promotes accessible and culturally aware emotional well-being for all. Key points include:

Combating Bias: JournalBot utilizes diverse training data to mitigate bias. By understanding diverse communication styles and complex emotions, JournalBot fosters a safe space for open expression, regardless of a user's background.

Building Belonging: Building a sense of belonging by fostering open communication about youth’s well-being.

Accessible and Confidential: Free and web-based, JournalBot tackles the stigma and removes cost and location limitations.

Promoting Self-Awareness: It empowers all to manage their emotional well-being through self-discovery, exploration of feelings, and mood tracking.

How we built it

We leveraged Cohere's powerful API to create the conversational AI core, enabling JournalBot to respond and guide users. Next, we curated a diverse dataset to train our sentiment analysis model, ensuring it understands emotional nuances across diverse communication styles. Finally, we utilized React to build a user-friendly and accessible web interface, providing a safe space for young adults to explore their emotional well-being.

Challenges we ran into

Embracing API Deployment: Our team faced the challenge of API deployment, turning it into an opportunity for skill growth. We sought mentor guidance and dove into API documentation and online forums to conquer this learning curve.

Frontend-Backend Fusion: Integrating frontend and backend components posed challenges, but our team embraced the opportunity to refine our skillset!

Accomplishments that we're proud of

We're proud of contributing to tackling the critical issue of mental health accessibility for young adults. We meticulously curated a diverse dataset to train our sentiment analysis model. This, combined with personalizing Cohere's API prompts with personalized preambles, empowers JournalBot to provide insightful prompts and analyze user reflections, creating a safe space for self-exploration.

What we learned

Our most significant takeaway is the importance of embracing nuance. Specifically, in prompt engineering, using precise and concise language during training and modifying Cohere's API proved indispensable for achieving accurate sentiment analysis and delivering a personalized, compassionate chatbot experience.

What's next for Journal-Bot

Empowering Through Language: While currently English-focused, future plans include multilingual capabilities to break down language barriers to increase accessibility

Beyond Text: We envision incorporating voice and image recognition features. This will cater to users with different learning styles and preferences, further personalizing the self-exploration experience.

Expanding the Support System: In the future, we aim to integrate with mental health resources and professionals. This could involve connecting users with relevant hotlines, therapists, or support groups based on their needs. Partnering with mental health organizations would strengthen JournalBot's role as a bridge to professional care.

Collecting Culturally Diverse Data: To ensure JournalBot's sentiment analysis accurately reflects the experiences of minority groups and different cultures, we will actively collect and integrate data from diverse user bases. This will involve partnerships with organizations serving these communities and ongoing efforts to broaden our reach. A more inclusive dataset will lead to fairer and more culturally sensitive responses

Built With

• cohere
• flask
• html/css
• javascript
• python
• react