mloss.org Scalable Parallel EM Algorithms for LDA in Multi Core Systemshttp://mloss.orgUpdates and additions to Scalable Parallel EM Algorithms for LDA in Multi Core SystemsenFri, 05 Dec 2014 10:18:36 -0000Scalable Parallel EM Algorithms for LDA in Multi Core Systems 1.0.0http://mloss.org/software/view/584/<html><p>Latent Dirichlet allocation (LDA) is a widely-used probabilistic
topic modeling tool for content analysis such as web mining. To
handle web-scale content analysis on just a single PC, we propose
multi-core parallel expectation-maximization (PEM) algorithms to
infer and estimate LDA parameters in shared memory systems.
By avoiding memory access conflicts, reducing the locking time
among multiple threads and residual-based dynamic scheduling,
we show that PEM algorithms are more scalable and accurate than
the current state-of-the-art parallel LDA algorithms on a commodity
PC.
</p></html>Xiaosheng Liu, Jia Zeng, Xi Yang, Jianfeng Yan, Qiang YangFri, 05 Dec 2014 10:18:36 -0000http://mloss.org/software/rss/comments/584http://mloss.org/software/view/584/ldaparallelem