Merge pull request #477 from superlinked/update-article---recsys-basic

Update recSys-basic.md

docs/articles/recSys-basic.md

@@ -6,8 +6,8 @@ Recommender Systems are central to nearly every web platform that offers things
 
 Below, we'll show you how to build a very simple recommender system. The rationale for our RecSys comes from our general recipe for providing recommendations, which is based on user-type (activity level):
 
-| **interaction level** | -> | **recommendation approach** |
-| ----------------- | -- | ---------------- |
+| **interaction level** | -> &nsbp; | **recommendation approach** |
+| :----------------- | :-- | :---------------- |
 | no interactions (cold start) | -> | most popular items |
 | some interactions | -> | content-based items |
 | more interactions | -> | collaborative filtering (interaction-based) items |