Skip to content

This is a single-page application (SPA) developed using Nuxt 3.2.2 and FastAPI 0.92.0 for predicting the anxiety probability of a given tweet (or any text prompt limited to max. 280 characters).

License

Notifications You must be signed in to change notification settings

mdrhmn/tweets-anxiety-detector-v2

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

89 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Tweets Anxiety Detector

This is a single-page application (SPA) developed using Nuxt 3.2.2 and FastAPI 0.92.0 for predicting the anxiety probability of a given tweet (or any text prompt limited to max. 280 characters).

Disclaimer: Given the extremely small dataset that the ML model is trained with, the prediction may not be accurate/reliable.


Introduction

This app is a mini side project related to my award-winning final year project (FYP) titled Fitweet: Arduino-based Smart Watch for Early Anticipatory Anxiety Notification System.

Our binary classification Emotion Detection Model (EDM)—a Logistic Regression model trained with 4,700 bilingual English and Malay tweets from Malaysian university students aged 18 and 25 inclusive—is used to analyse the anxiety of the tweet.

The emotion prediction will be analysed into 2 emotions - Happy or Worry. The emotion detected is considered as No Emotion if the probabilities of Happy and Worry emotions are contentious.

To display the result, we used a tool called LIME. LIME is a technique that approximates any black box machine learning model with a local, interpretable model to explain each individual prediction.


Installation

Build the images and spin up the containers:

$ docker-compose up -d --build

Ensure http://localhost:8000, http://localhost:8000/docs, and http://localhost:5000 work as expected.


References

  1. Heroku. (n.d.). Building Docker Images with heroku.yml [Documentation]. Retrieved from https://devcenter.heroku.com/articles/build-docker-images-heroku-yml
  2. Stack Overflow. (2016, June 23). Docker error: bind address already in use [Online forum post]. Retrieved from https://stackoverflow.com/questions/37971961/docker-error-bind-address-already-in-use
  3. TestDriven.io. (2021, March 15). Developing a Single-Page App with FastAPI and Vue.js [Blog post]. Retrieved from https://testdriven.io/blog/developing-a-single-page-app-with-fastapi-and-vuejs/
  4. TestDrivenIO. (n.d.). fastapi-vue [GitHub repository]. Retrieved from https://github.com/testdrivenio/fastapi-vue/edit/main/README.md
  5. Tiangolo, S. (n.d.). FastAPI Tutorial [Webpage]. Retrieved from https://fastapi.tiangolo.com/tutorial/
  6. Zhang, D. (2019, January 8). Heroku + Docker in 10 Minutes [Blog post]. Towards Data Science. Retrieved from https://towardsdatascience.com/heroku-docker-in-10-minutes-f4329c4fd72f#ee9e

About

This is a single-page application (SPA) developed using Nuxt 3.2.2 and FastAPI 0.92.0 for predicting the anxiety probability of a given tweet (or any text prompt limited to max. 280 characters).

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published