• 

  Sign up page

• 

  Dashboard

• 

  Product Ingredients Analysis

• 

  Flowchart

Healthify- Calgary Hacks 2022

API Blueprint: https://docs.google.com/document/d/1UibA3Kuea-CXpcYSPyptzJmxg-iko_Sk0zLF-7Iy6L0/edit?usp=sharing

API postman document: https://documenter.getpostman.com/view/18814228/UVkjwJFs

What is Healthify ?

Healthify is a one stop personalised platform that simplifies the concept of living a health lifestyle. Healthify platform employs a machine learning model that generates videos, research papers, visualized data as a response to the users input. The Project stems from our personal hardships of maintaining a healthy lifestyle. It is a mobile friendly web application which serves the needs of a user based on their input. The platform is split into two main components.

1. Feature to verify authenticity of health related subjects
   
2. Feature to identify if a product is suitable for the users use.
   • product is identified using a barcode scanner
     

Challenges we ran into:

We ran into multiple challenges along the development phase and were dealt with accordingly. With respect to Front-End we had challenges with aligning of placeholders that were used to display the results from our Machine Learning model. With respect to Back-End we ran into time bounded challenges as Machine learning model takes a considerable amount of time to create the tensor file. Employing a Machine Learning model also posed as challenges as the packages to be downloaded were conflicting with existing packages, resolving dependencies also required a lot of time dedication. Our API choice was another thing we struggled with as we wanted to ensure we were using the best and more reliable technologies with respect to development and debugging.

What we learnt:

A sprint coding competition always teaches us how to create valuable code within a restricted time frame. We came across multiple new technologies and discovered numerous ways to integrate all the segments of the project to provide a well working web application, Furthermore while creating a dataset for our 2nd feature we learnt about various harmful ingredients in products.

Technologies Used

API's used:

• Google Cloud Vision API
• IBM Watson speech to text API
• Youtube API
• Wikipedia API
• Open Food Facts API

Integration of frameworks

Healthify web application was developed as a React app. Available to be viewed on mobile as well. The Back-End aspect of the code consists of a Machine Learning Model, MySQL database, and Flask API. Machine Learning model uses non textual input and provides relevant youtube suggestions, research papers, and more. Healthify is personalized for each user. The data pertaining to each user is stored in a MySQL database. The database also contains the search history of each user from previous interactions with the platform. Flask API was used to act as a controller between the Machine Learning model, Database and Front-End

Dataset used

Reference : https://www.kaggle.com/allen-institute-for-ai/CORD-19-research-challenge?select=metadata.csv COVID-19 Open Research Dataset (CORD-19). CORD-19 is a resource of over 500,000 scholarly articles, including over 200,000 with full text, about COVID-19, SARS-CoV-2, and related coronaviruses. This dataset provides information about several misconceptions known to people. This dataset assists people around to world to navigate through solutions to their misconceptions.

ML model integration

• The model is built using allenai-specter model with SentenceTransformers.
• The dataset is fed to the model with the relevant information (abstract) required to recommend articles that solves the misconception.
• Top article is send as an output with the url and abstract of the Research article with the similarity score.

External Contribution

Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change. Please base all pull requests off of the main branch as they will be rebase merged. Check that there are no linter issues before opening a pull request by running.

Prequisites to run the ML Model

pip install regex
pip install torch
pip install pickle5
pip install sentence-transformers
pip install pandas
pip install termcolor

License

Built With

• bert
• css3
• google-cloud
• google-vision-api
• html5
• ibm-watson-speech-to-text-api
• javascript
• jupyternotebook
• machine-learning
• mysql
• open-food-facts-api
• python
• react