update doc

index.rst

@@ -7,14 +7,14 @@ Get Started with xLearn !
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
 
 xLearn is a *high-performance*, *easy-to-use*, and *scalable* machine learning package, 
-which can be used to solve large-scale machine learning problems, especially for for solving 
+which can be used to solve large-scale machine learning problems, especially for for the 
 problems on large-scale sparse data, which is very common in scenes like CTR prediction and 
 recommender system. If you are the user of liblinear, libfm, or libffm, now xLearn the your 
-another better choice. This is because xLearn handles all of these models in a uniform platform 
-and provides better performance and scalability.
+another better choice. This is because xLearn handles all of these models in an uniform 
+platform and provides better performance and scalability.
 
-This is a quick start tutorial showing snippets for you to quickly try out xLearn on a small demo 
-dataset (Criteo CTR) for a binary classfication task.
+This is a quick start tutorial showing snippets for you to quickly try out xLearn on a small 
+demo dataset (Criteo CTR prediction) for a binary classfication task.
 
 Link to the Other Helpful Resources
 ----------------------------------------