1 / 17

Neighborhood Based Fast Graph Search in Large Networks

Neighborhood Based Fast Graph Search in Large Networks. Arijit Khan Nan Li Xifeng Yan Ziyu Guan Supriyo Chakraborty Shu Tao SIGMOD 2011. Outlines. Motivation Objectives Methodology - Ness -Information Propagation Model Experiments Conclusions.

geoff
Download Presentation

Neighborhood Based Fast Graph Search in Large Networks

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Neighborhood Based Fast Graph Search in Large Networks ArijitKhan Nan Li Xifeng Yan Ziyu Guan SupriyoChakrabortyShuTao SIGMOD 2011

  2. Outlines • Motivation • Objectives • Methodology -Ness -Information Propagation Model • Experiments • Conclusions

  3. Motivation • Entity-relationship graphs and social networks are very large and complex with a lot of attributes associated. • It is hard to come up with a query that exactly conforms with the graph structures in the target network due to the lack of schemas in linked data.

  4. Objectives • As long as the the proximity between these entities is approximately maintained in a query graph, shall be consider matches. • propose graph similarity search framework to determine approximate matches in massive graphs.

  5. Example: Top-1 Match for Query

  6. Methodology • Ness (Neighborhood Based Similarity Search). • :

  7. Embedding • Embedding written as .

  8. Cost function • Edge mismatch cost function

  9. Cont. • Embedding is a better match than

  10. Information Propagation Model • U’s neighbors is propagated to u through different paths and accumulated at u. • Convert each node into a multidimensional vector.

  11. Neighborhood-based Cost Function

  12. Cost of Embedding

  13. Cost of Embedding

  14. Experiments

  15. Conclusions • Empirical results show that it could quickly and accurately find high-quality matches in large networks, with negligible time cost. • In future work, it will be interesting to consider the graph alignment problem, when the node labels in two graphs are not exactly identical.

More Related