Create README.md

README.md

@@ -0,0 +1,39 @@
+---
+language:
+- en
+dataset_info:
+  features:
+  - name: labels
+    sequence:
+      class_label:
+        names:
+          0: negative
+          1: neutral
+          2: positive
+  splits:
+  - name: test
+    num_examples: 785
+  - name: train
+    num_examples: 4446
+annotations_creators:
+- expert-generated
+license:
+- mit
+multilinguality:
+- monolingual
+pretty_name: financial_news_sentiment_mixte_with_phrasebank_75
+size_categories:
+- 1K<n<10K
+source_datasets:
+- original
+tags: []
+task_categories:
+- text-classification
+task_ids:
+- multi-class-classification
+- sentiment-classification
+---
+# Dataset Card for "financial_news_sentiment_mixte_with_phrasebank_75"
+
+This is a customized version of the phrasebank dataset in which I kept only sentence validated by at least 75% annotators.  
+In addition I added ~2000 articles validated manually on Canadian financial news.