LDA_GPU
Project Information
- Discipline
- Computer Science (401)
- Orientation
- Research
This project investigates a scalable GPU-based solution for author-topic model LDA algorithm. This work will be applied to publication data and analyzing the hidden topic model related to documents, words and authors.
Intellectual MeritThe outcome will support analyzing patents or domain-specific corpus.
Broader ImpactsThe result will provide data analysis service to the library science community.
Project Contact
- Project Lead
- Zhenghao Gu (guzh)
- Project Manager
- Judy Qiu (xqiu)
Resource Requirements
- Hardware System
-
- delta (GPU Cloud)
To implement and test a LDA_GPU algorithm.
Scale of Use2 - 4 GPU nodes
Project Timeline
- Submitted
- 07/02/2014 - 15:28