Change LICENSE and add note

README.md

@@ -1,5 +1,5 @@
 ---
-license: mit
+license: cc-by-nc-4.0
 task_categories:
   - graph-ml
 size_categories:
@@ -25,7 +25,7 @@ We further subsampled and cleaned the data to focus on application download inte
 The dataset has two main files.
 
 - `myket.csv`: This file contains the interaction information and follows the same format as the datasets used in the "[JODIE: Predicting Dynamic Embedding Trajectory in Temporal Interaction Networks](https://github.com/claws-lab/jodie)" (ACM SIGKDD 2019) project. However, this data does not contain state labels and interaction features, resulting in associated columns being all zero.
-- `app_info_sample.csv`: This file comprises features associated with applications present in the sample. For each individual application, information such as the approximate number of installs, average rating, count of ratings, and category are included. These features provide insights into the applications present in the dataset.
+- `app_info_sample.csv`: This file comprises features associated with applications present in the sample. For each individual application, information such as the approximate number of installs, average rating, count of ratings, and category are included. These features provide insights into the applications present in the dataset. **Note**: The number of installs, average rating, and the count of ratings info are gathered during the period of interaction data collection, so they may cause information leakage if used for interaction prediction tasks.
 
 ## Dataset Details