Update README.md

This is the Turkish SMS spam dataset employed in the following study.
Uysal, A. K., Gunal, S., Ergin, S., & Gunal, E. S. (2013). The impact of feature extraction and selection on SMS spam filtering. Elektronika ir Elektrotechnika, 19(5), 67-72.

I removed the labels (legitimate, spam) because of using it for text generation task.

README.md

@@ -13,7 +13,11 @@ dataset_info:
     num_bytes: 9681.4
     num_examples: 85
   download_size: 59105
-  dataset_size: 96814.0
+  dataset_size: 96814
+task_categories:
+- text-generation
+language:
+- tr
 ---
 # Dataset Card for "turkishSMS-ds"