Updated README

README.md

@@ -9,25 +9,31 @@ All datasets are sub-grouped into letter directories based on the
 initial letter of the dataset name.  Within each letter directory is a
 directory for datasets that contain labels.  Within each dataset
 directory is a subdirectory from the creator of the labels, with
-"provider" being a reserved subdirectory name for labels that came
-from the dataset provider.
+"provider-somenamehere" being a reserved subdirectory name for labels that came
+from the dataset provider. 
 
 As an example, labels created by the provider for the
 /B_Root_Anomaly-20190907/ dataset can be found in:
 
-    B/B_Root_Anomaly-20190907/provider
-
+    B/B_Root_Anomaly-20190907/provider-uscisiant
+
+When citing labels, please use the label name (omit "provider" if present)
+and dataset name. For example provider-uscisiant labels from the above
+example could be cited as "uscisiant labels for B_Root_Anomaly-20190907
+dataset".
+
 In each subdirectory containing labels, please refer to the specific
 README file that describes the associated labels or tools to create
 the labels and how to use them.
 
-Each leaf folder relates to a specific dataset
-and contains labeling program if any, and information needed to
+Each leaf folder relates to a specific dataset. Labels come in many
+flavors - per-record or per-event labels. In some cases there will be 
+a labeling program or pointers to it, and information needed to
 run this program with the original dataset to obtain per-record labels.
 
 ## Supporting tools
 
-In some cases, tools have created to "generate" labels using the
+In some cases, tools have been created to "generate" labels using the
 dataset as an input.  These tools, and descriptions about how to
 compile and use them, will either be in the dataset subdirectory
 itself or may exist in the /tools/ subdirectory, under the label