diff --git a/data_science.md b/data_science.md
index cc2e1a4..f690c11 100644
--- a/data_science.md
+++ b/data_science.md
@@ -1,15 +1,18 @@
 Python - defacto for data science.
 
-What do we need:
+What do we need
+---
 Tabular Data Framework
 Charting support
 Linear algebra
 Statistics
 ML
 
-What are the alternatives 
+What are the alternatives to Python/R/Julia/Matlab
+---
 
 Java:
+---
 
 Options:
 JTableSaw
@@ -34,7 +37,8 @@ http://beakerx.com/
 https://github.com/twosigma/beakerx
 
 
-Scala:
+Scala
+---
 https://github.com/ThoughtWorksInc/DeepLearning.scala
 Uses Java nd4j underneath - provided by Thoughtworks
 http://beakerx.com/ - also provides Scala
@@ -43,8 +47,13 @@ http://beakerx.com/ - also provides Scala
 https://typelevel.org/spire/
 Spire is a numeric library for Scala which is intended to be generic, fast, and precise.
 
+Spark eco-system:
+https://spark.apache.org/mllib/
+https://github.com/apache/spark/tree/master/mllib/src/main/scala/org/apache/spark/mllib/linalg
 
-c++:
+
+c++
+---
 http://arma.sourceforge.net/ Armadillo - C++ library for linear algebra & scientific computing
 https://gitlab.com/conradsnicta/armadillo-code
 https://www.mlpack.org - mlpack - fast, flexible C++ machine learning library