Suicide

CohereAIHack

Documentation

Problem Formulation

AISaturdaysIbadan

Amrs/ ai-content-creator - app

Problem Formulation

A lot of people use the Social Media and put out their thought there , most of the time we Hear of people committing suicide after posting something very suicidal but this doesnt start In a day. Their posts and activities must have been showing this trend.

Giving the whole focus on mental health lately , nothing stops the social media Network from contributing their own quota to solving mental health issues.

With the vast amount of data generated on these social media networks , nothing Stops its from being used to help people have better mental health.

A useful tool for this is sentiment analysis.Instead of using sentiment analysis just to check for offensive posts and whether people love a product, the emotion and subjective feelings contained in posts on the social networks are analysed and used to predict the trends of actions of individuals. With this, the social media network can use machine learning algorithms to predict the physical actions of people based on their online presence and trends in their posts.Also, they can use this information to help healthcare professionals to know more about individuals and help them in a better way, leading to better personalised mental healthcare for the people. The social media can then use generative ai to generate responses that match the person's emotions and trends in online activity.

In this way, social media is not just going to be for posts and generating money but is going to be a place where the health of the users actually matter and is treated with optimum priority for the betterment of the masses and the healthcare system as a whole.

Implementation of Idea and Mechanism of Action. Using Cohere’s API makes everything very easy. We analyse our text and classify it using cohere’s classifier api. Then the result of our analysis will be used by gpt api to generate a text that matches the post used and uses it to comfort, encourage, etc the individual.

Psychologists and mental health advocates can also benefit from this a lot by this serving as a link between the people and the healthcare professionals. Hence with our solution tagged inner health, we move the war against mental health illnesses from the physical to the online platforms, hence fighting on two fronts.

We built our project using the Python language and then we used cohere APIs for classification and language generation API. We analyzed the tweets for 6 different emotions and then based on the emotions and the post generate a text that advises, encourages, or lightens them up as the case may be.

Reference https://www.hindawi.com/journals/cin/2022/9194031/ Others

What's next for Internal Health

Built With

cohere
colab
github
python
render
streamlit

Submitted to

#CohereAIHack: Empower your Creativity with Cohere's Multilingual Embeddings
- Winner Honourable mentions

Created by

I worked on training the model using a labeled twitter dataset of emotions. I cleaned the data and randomised it, picking 50 as required by the cohere classifier api to classify posts based on emotions.

I then used the cohere generative api to generate responses based on the post and the emotions in the post. This serves to augment and help the person have better emotional states. Also, there's a visualizer to show the analysis of post and emotional contents of the posts. That's my part. The machine learning encoding aspect.

Halleluyah Oludele
Medical Student | Data Science/Machine Learning | Medicine+ Tech| Automation| Python| Member of AISaturdaysIbadan
Right from brainstorming ideas to integration and problem formulation, hacking on the hardware has always been a crucial part of the process. After that, I set up the GitHub repository and serve as the project manager to ensure that everything is in order. I also make sure that everything works fine and include a brief documentation for people to understand the whole process.

I am excited to contribute to the team and wish us success.

Qudus Aileru
Data and Information Science Student at University of Ibadan, Ibadan, Nigeria || A member of AISaturdaysIbadan Community.
I had the opportunity to work with Streamlit for the first time, and it was an exciting experience. My main task was to integrate the backend functionality created by my colleagues and render the results on the front end using Streamlit. I found the platform to be user-friendly, and I was able to quickly learn how to use it to achieve the desired results.

I also used Render to deploy our application, and it was a straightforward process. Our application is now available online at internal-health.onrender.com, and I am proud to have played a role in its creation.

This project was a great learning experience for me, and I gained valuable new skills that I can apply to future projects. I also appreciate the opportunity to contribute to a project focused on mental health, which is a cause that is important to me.

Timothy Adeleke