Cyberify | Devpost

Home Page

Inspiration

Nowadays everyone around the world is using a website/apps in a significant amount for daily tasks. Because of heavy usage of various apps created by various developers/organizations, we may encounter suspicious services that pose a risk to our data but even for reputed companies, it is important to know what data is being used. Nowadays, it is practically impossible for most users to read the privacy policy, which consists of thousands of words. It would be much more convenient and easier to use if we could make it short and summarize key points. That's how we came up with the idea for Cyberify.

What it does

It takes full privacy policy in the form of text and summarizes it shortly along with a few key points like "Cookies", "Data Around World", "Personal Data Storing", etc.

How we built it

Front-End: HTML CSS Javascript Back-End: Node JS + Express, Flask ML: Python + Gensim + Sentence Transformers

Our Features: Summary: Using gensim Library Does the website Store Personal Data: We calculated it using NLP. We first generated strings of max length 3 from our text and then calculated the cosine similarity of this with the string “use personal data” and if the cosine similarity is greater than our threshold value then we say the personal data is stored and used else we conclude that it is not stored.

Cookies: We calculated it using NLP. We first generated strings of max length 3 from our text and then calculated the cosine similarity of this with the string “cookies are used” and if the cosine similarity is greater than our threshold value then we say the cookies are used else we conclude the cookies are not used. Data Around the world, Do track you: Also uses the same method as above.

ReadScore: This tells us how easy it is to read the content and we calculated it using textstat. SmogScore: The SMOG grade is a measure of readability that estimates the years of education needed to understand a piece of writing. SMOG is an acronym for "Simple Measure of Gobbledygook". We build it using textStat library Sentiment: We calculated it using the texBlob library.

Challenges we ran into

Our first main challenge emerged during the ideation phase. We wanted to solve a unique problem. Thus, brainstorming with different tracks for different problems was a challenge while working remotely. The second challenge was the tech implementation. Our goal was to make an interactive user interface with accurate summarizer and to integrate it using APIs. Therefore, planning and implementing the flow were a bit challenging.

Accomplishments that we're proud of

We overcame all of these challenges, and I am proud to announce that we have developed V1.0 of Cyberify

What we learned

Learned to work around Machine Learning algorithms/models and its deployment.

What's next for Cyberify

In the coming days, we plan to add more features that our target users want. For instance, we want to develop some kind of browser extension that can be triggered when visiting the privacy policy page. Also, planning to improve our summarizer algorithm.