README

yiylin · Oct 16, 2015 · 26c22f9 · 26c22f9
1 parent a33a0d3
commit 26c22f9
Showing 1 changed file with 3 additions and 1 deletion.
diff --git a/README.md b/README.md
@@ -2,7 +2,9 @@
 
 LightLDA is a distributed system for large scale topic modeling. It implements a distributed sampler that enables very large data sizes and models. LightLDA improves sampling throughput and convergence speed via a fast O(1) metropolis-Hastings algorithm, and allows small cluster to tackle very large data and model sizes through model scheduling and data parallelism architecture. LightLDA is implemented with C++ for performance consideration.
 
-We have sucessfully trained big topic models (with trillions of parameters) on big data (Top 10% PageRank values of Bing indexed page, containing billions of documents) in Microsoft. For more technical details, please refer to our [WWW'15 paper](http://www.www2015.it/documents/proceedings/proceedings/p1351.pdf). For documents, please view website [http://www.dmtk.io](http://www.dmtk.io).
+We have sucessfully trained big topic models (with trillions of parameters) on big data (Top 10% PageRank values of Bing indexed page, containing billions of documents) in Microsoft. For more technical details, please refer to our [WWW'15 paper](http://www.www2015.it/documents/proceedings/proceedings/p1351.pdf). 
+
+For documents, please view our website [http://www.dmtk.io](http://www.dmtk.io).
 
 ##Why LightLDA