Update README.md

README.md

@@ -9,10 +9,16 @@ size_categories:
 - 100M<n<1B
 ---
 
+- **Homepage:** [Dell Research Harvard](https://dell-research-harvard.github.io/)
+- <!--- **Repository:** [GitHub repository]()--->
+- <!--- **Paper:** [NeurIPS Submission]()--->
+- **Point of Contact:** Melissa Dell
+
+
 ![headline_image_with_title.jpg](https://s3.amazonaws.com/moonup/production/uploads/61654589b5ec555e8e9c203a/fJiAy43PFD3FBa1aMSzHz.jpeg)
 
 ## Dataset Summary
-HEADLINES contains 396,001,930 pairs of different headlines for the same newspaper article. 
+HEADLINES is a massive semantic similarity dataset, containing 396,001,930 pairs of different headlines for the same newspaper article, taken from historical U.S. newspapers, covering the period 1920-1989. 
 
 ## Dataset Creation
 The dataset was constructed using a large corpus of newly digitized articles from off-copyright, local U.S. newspapers. 
@@ -24,7 +30,7 @@ The dataset contains 396,001,930 positive semantic similarity pairs, from 1920 t
 
 ![image (3).png](https://s3.amazonaws.com/moonup/production/uploads/61654589b5ec555e8e9c203a/vKeR-SEEfYte6ZZbdpaq3.png)
 
-It contains headlines from all 50 states
+It contains headlines from all 50 states. 
 
 ![map (1).png](https://s3.amazonaws.com/moonup/production/uploads/61654589b5ec555e8e9c203a/0WMdO8Fo1nfYiId4SlWaL.png)
 
@@ -39,4 +45,4 @@ It contains headlines from all 50 states
 
 
 ## Contacts
-
+This dataset was created by Emily Silcock and Melissa Dell. For more information, see https://dell-research-harvard.github.io/