Giter Site home page Giter Site logo

Comments (8)

vigna avatar vigna commented on August 27, 2024

You can certainly use the framework for that albeit an EFGraph might be more appropriate. We never worked on that tho...

from webgraph-big.

cfangplus avatar cfangplus commented on August 27, 2024

We found that many papers cites the layered label propagation paper, among which includes graph processing system like GraphX/Gemini/PowerGraph and graph neural network system like P3. Specially, many of them use dataset like uk-2007-05. However, we found those who opened their source code like GraphX/Gemini system tried to transfer the .graph file into <source id, target id> txt format via ArcListASCIIGraph class. After this operation, the graph data loaded into memory is not compressed. So when the graph data is large, there may cause out of memory issue, but we thought what WebGraph/LLP really aims to do is to resolve this memory issue when facing huge large graph. That's really confused me, do you have any suggestion?

from webgraph-big.

vigna avatar vigna commented on August 27, 2024

We distribute data in the WebGraph format, and a lot of people use the data, but the same people does not want to use the WebGraph interface, so they convert to ASCII. Sad, but it works for them apparently.

from webgraph-big.

cfangplus avatar cfangplus commented on August 27, 2024

We also observed an experiment that we list the figure as follow and it's from the paper of graph processing system Gemini.
Hereby they set up an experiment for their called chunk-based partitioning method which partition graph with load balance. Although they use the <source id, target id> txt format rather than .graph format, the results show that locality property of uk-2007-05 graph data is very good which helps getting a better performance. So I think although graph data with <source id, target id> txt format is not compressed but the node ordering still helped to maintain the locality property, right? Sorry I have not totally understand the theory of WebGraph or LLP, so need your help to check the judgement.
捕获

from webgraph-big.

vigna avatar vigna commented on August 27, 2024

Totally. The ordering of the graph is a form of clustering. Chunks of consecutive nodes have dense connections.

from webgraph-big.

cfangplus avatar cfangplus commented on August 27, 2024

Great thx.

from webgraph-big.

cfangplus avatar cfangplus commented on August 27, 2024

Just one idea, I observed that BVGraph/EVGraph/ASCIIGraph class provide many load and store methods, I think maybe WebGraph could provide more methods for fast random access to a link , such as fast access to neighbors or subgraph for one node. If so, maybe people would more likely trends to use the WebGraph interface rather than convert to ASCII to get coo data format which is not compressed.

from webgraph-big.

vigna avatar vigna commented on August 27, 2024

I think you really misunderstood what WebGraph does. Please have a look at ImmutableGraph.

from webgraph-big.

Related Issues (1)

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.