In today's world, text summarization has become ubiquitous. But many times the user wants a very short way to summarize a big document. SumViz tries to solve this problem by creating a graph-based, easy to read and context full summarization.
As we all projects, we ran into many problems. Firstly, was figuring out how data will pipeline from the python ML modules to the D3.js based graph. Since the data is parsed into a very different looking schema when compared to how the JSON data schema of our D3.js plot. Further, figuring out how to filter for the most relevant entity-relationship map was quite challenging for us. But in the end, we tried our best and did come up with good and functioning solutions to the above-mentioned problems.
Discussion