Skip to content

ByteSize allows users to find information in documents with phrases and related terms as opposed to Command-F's requirement of being verbatim. This project was made during UMD's Bitcamp Hackathon competititon.

Notifications You must be signed in to change notification settings

arjunaggarwal03/bytesize

Repository files navigation

ByteSize

Summary

ByteSize allows users to find information in documents with phrases and related terms as opposed to Command-F's requirement of being verbatim. This project was made during UMD's Bitcamp Hackathon competititon.

Credit for model: Semantic Similarity model

Project Design

Design

Credit for diagram: GitDiagram

Demo

Website Image

In the image above the user's text is located in the top left element and the paraphrased search key is in the element below. The results of the model are displayed on the right. Within each result item, you can see the relevant sentence and the score given by the model.

Places to Improve

In some cases, the sentences proposed by the model are irrelevant to the search key. Furthermore, the UI has the following bugs: rendering issues with repeated searches & issues handling multiple search keys.

About

ByteSize allows users to find information in documents with phrases and related terms as opposed to Command-F's requirement of being verbatim. This project was made during UMD's Bitcamp Hackathon competititon.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published