Is this my friend?

A web application that analyzes your friend's Hacker News comments and posts and classifies if it really is them based on sentiment analysis.

Comment

Inspiration

With so many accounts being compromised in data breaches every year within the social media space, its hard to keep track whether a person posting something online is really them.

Our concept surrounds the question, "Is there a way I can tell if this person is really the same person based on their past social media and current social media activity?"

For our project we have chosen to analyze user's Hacker News' posts but plan to expand to other social media platforms.

What it does

"Is this my friend?" analyzes a user's hacker news' posts and comments to generate the sentiment displayed in them. In the future, we plan to compare the sentiment from the current month with the user's past statistics and generate a score to check whether user is the same person or chance of anomalous behavior.

How we built it

Brainstorming

We spent the first few hours to brainstorm among the different project categories before finalizing on the "Ever Vigilant" Hack category.

Design

Following deciding the idea we all worked on different aspects of it, the initial figma designs, created by Zi were really helpful to guide our frontend development at a later stage

Collaboration

We used a github repository and features such as branches, pull requests and merged any new features we worked on. When one of the team members was stuck on an aspect, we usually got on a slack call to debug it with them.

Accessibility First

When designing interfaces, we tried our best to follow some rules from WCAG. For example, all the color we used passed the test from Colour Contrast Analyzer (CCA) and achieved a result of AAA level. Also, when coding the program, we added description for meaningful images as well as leave the description blank for all decorative images.

Challenges we ran into

Our initial idea was to analyze user's Twitter tweets, but the Twitter developer API request process was slightly longer and was still under processing, so we decided to choose an alternate API.

Accomplishments that we're proud of

Base concept and MVP . When designing interfaces, we tried our best to follow some rules from WCAG. For example, all the color we used passed the test from Colour Contrast Analyzer (CCA) and achieved a result of AAA level. Also, when coding the program, we added description for meaningful images as well as leave the description blank for all decorative images.

What we learned

  • Working with various APIs - Hacker News, Azure Sentiment Analysis API
  • Accessibility guidelines for the Web
  • Learnt and experimented with React

What's next for Is this my friend?

  • Incorporate functionality to analyzing Reddit user's posts
  • Incorporate functionality to analyze Twitter user's tweets
  • Incorporate code with more semantic elements and look more into how transition animation could probably influence usability

Built With

Share this project:

Updates