Git Product home page Git Product logo

Comments (4)

ankit avatar ankit commented on June 24, 2024

1 - Any other ideas on how we could make the second view better?

  • Do you think ability to bookmark/save biclusters for later reference will be useful?
  • Or maybe a history of each bicluster you have explored in a session? There could be a History Link which pops up all the previous biclusters you have explored. You can then select one to explore a bicluster again.

2 - might be a lower priority though (not sure). Makes sense to be able to reduce the noise.

3 - makes sense i think

4 - Do you care which bicluster the entities belong to? As a user, aren't you mostly concerned about the relationships between entities, rather than identifying them with their biclusters?

Thanks for the feedback! Very useful!

from see.

JacobMoore avatar JacobMoore commented on June 24, 2024
  1. I think the ability to bookmark biclusters could be useful, but we need to solve other problems first. I wasn't finding anything I could string together as a story so I had no desire to bookmark things. I was also using paper and the documents a lot so that is where I did most of my history and stringing things together. I think the history is a low priority.
  2. I think this is important, though I'm not sure how feasible it is. The terms "FBI", "CIA" and "NSA" seemed to link everything together and gave a lot of false positives in terms of groups. These were just the agencies that gathered the intelligence. If we could get rid of these elements, I think the data would be a lot sparcer and a lot more helpful.
  3. No additional comment
  4. I had trouble telling what were the new elements when hovering around, particularly for smaller biclusters. If I focus on a big bicluster to start with, that dominates the display. I wanted to know what the other biclusters added to the elements.

from see.

JacobMoore avatar JacobMoore commented on June 24, 2024

In terms of being able to make a story from the data, I'm not sure how feasible it is with biclusters now. The story had some tenuous connections that were still very important. I don't know of a way to filter these important but small connections from the unimportant and small connections other than reading the documents themselves.

We may need to just leave the story telling up to the user. Our tool is helpful for searching. Maybe we instead focus the second view on being a single search. Enter an element and the tool displays all biclusters that contain that element? Along with a list of all the elements in these biclusters? Its a big change for late in the project though.

from see.

ankit avatar ankit commented on June 24, 2024

I think the most you can do with biclusters is to come up with a group of entities you think are related and the most important. You still need textual content for knowing the relationships between these entities. That's why JigSaw was useful, it had a Documents View. We don't have access to the textual content, and I think developing a tool that links to the textual content is out of scope for a term project. Plus, we didn't get access to any text anyways in the first place.

I think we can keep the second view. I am building the ability to search biclusters based on entity name.

About the 4th point in your earlier comment, it's a double edged sword. You're right we're losing some data there, maybe highlight the newly added entities in some way, instead of highlighting the common ones.

from see.

Related Issues (9)

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.