Several documentation improvements to NSL.

- adds a link to the `examples` subdirectory from the main README.md file and from the TF website
- adds clarity to our contribution guidelines
- adds a README.md file under the `examples` subdirectory

PiperOrigin-RevId: 393784004

Author: arjung

README.md

@@ -71,38 +71,39 @@ learning with structured signals.
    target="_blank"><img src="http://img.youtube.com/vi/Js2WJkhdU7k/0.jpg"
                         alt="Adversarial Learning" width="180" border="2" /></a>
 
-We've also created the following hands-on colab-based tutorials that will allow
-you to interactively explore NSL:
+We've also created hands-on colab-based tutorials that will allow you to
+interactively explore NSL. Here are a few:
 
 *   [training with natural graphs](https://github.com/tensorflow/neural-structured-learning/blob/master/g3doc/tutorials/graph_keras_mlp_cora.ipynb)
 *   [training with synthesized graphs](https://github.com/tensorflow/neural-structured-learning/blob/master/g3doc/tutorials/graph_keras_lstm_imdb.ipynb)
 *   [adversarial learning](https://github.com/tensorflow/neural-structured-learning/blob/master/g3doc/tutorials/adversarial_keras_cnn_mnist.ipynb)
 
+You can find more examples and tutorials under the [examples](examples)
+directory.
+
 ## Contributing to NSL
 
 Contributions are welcome and highly appreciated - there are several ways to
 contribute to TF Neural Structured Learning:
 
-*   Case studies. If you are interested in applying NSL, consider wrapping up
+*   Case studies: If you are interested in applying NSL, consider wrapping up
     your usage as a tutorial, a new dataset, or an example model that others
-    could use for experiments and/or development.
+    could use for experiments and/or development. The [examples](examples)
+    directory could be a good destination for such contributions.
 
-*   Product excellence. If you are interested in improving NSL's product
+*   Product excellence: If you are interested in improving NSL's product
     excellence and developer experience, the best way is to clone this repo,
     make changes directly on the implementation in your local repo, and then
     send us pull request to integrate your changes.
 
-*   New algorithms. If you are interested in developing new algorithms for NSL,
+*   New algorithms: If you are interested in developing new algorithms for NSL,
     the best way is to study the implementations of NSL libraries, and to think
     of extensions to the existing implementation (or alternative approaches). If
     you have a proposal for a new algorithm, we recommend starting by staging
-    your project in the `research` directory and including a colab notebook to
-    showcase the new features.
-
-    If you develop new algorithms in your own repository, we are happy to
-    feature pointers to academic publications and/or repositories that use NSL,
-    on
-    [tensorflow.org/neural_structured_learning](http://www.tensorflow.org/neural_structured_learning).
+    your project in the [research](research) directory and including a colab
+    notebook to showcase the new features. If you develop new algorithms in your
+    own repository, we would be happy to feature pointers to academic
+    publications and/or repositories using NSL from this repository.
 
 Please be sure to review the [contribution guidelines](CONTRIBUTING.md).
 

g3doc/framework.md

@@ -75,9 +75,9 @@ NSL brings the following advantages:
 
 ## Step-by-step Tutorials
 
-To obtain hands-on experience with Neural Structured Learning, we have three
-tutorials that cover various scenarios where structured signals may be
-explicitly given, induced or constructed:
+To obtain hands-on experience with Neural Structured Learning, we have tutorials
+that cover various scenarios where structured signals may be explicitly given,
+constructed, or induced. Here are a few:
 
 *   [Graph regularization for document classification using natural graphs](tutorials/graph_keras_mlp_cora.ipynb).
     In this tutorial, we explore the use of graph regularization to classify
@@ -91,3 +91,7 @@ explicitly given, induced or constructed:
     In this tutorial, we explore the use of adversarial learning (where
     structured signals are induced) to classify images containing numeric
     digits.
+
+More examples and tutorials can be found in the
+[examples](https://github.com/tensorflow/neural-structured-learning/tree/master/neural_structured_learning/examples)
+directory of our GitHub repository.

neural_structured_learning/examples/README.md

@@ -0,0 +1,31 @@
+This directory contains examples and tutorials demonstrating how to use NSL.
+
+## Example Trainers
+
+The `.py` files in this directory are all example trainers runnable end-to-end
+as Python programs:
+
+*   `adv_keras_cnn_mnist.py`: Adversarial regularization on the MNIST dataset.
+*   `graph_keras_mlp_cora.py`: Graph regularization on the Cora dataset.
+*   `graph_nets_cora_gcn.py`: Graph Convolutional Network on the Cora dataset
+    using [GraphNets](https://github.com/deepmind/graph_nets).
+*   `graph_nets_cora_graph_regularization.py`: Graph regularization on the Cora
+    dataset using [GraphNets](https://github.com/deepmind/graph_nets).
+
+## Notebooks
+
+The
+[notebooks](https://github.com/tensorflow/neural-structured-learning/tree/master/neural_structured_learning/examples/notebooks)
+subdirectory contains colab-based tutorials that allow you to explore NSL
+interactively. This subdirectory is generally where most new tutorials are
+added. Note that the tutorials shown on our
+[TensorFlow website](https://github.com/tensorflow/neural-structured-learning)
+are hosted under the
+[g3doc/tutorials](https://github.com/tensorflow/neural-structured-learning/g3doc/tutorials)
+directory.
+
+## Data Preprocessing Scripts
+
+The
+[preprocess](https://github.com/tensorflow/neural-structured-learning/tree/master/neural_structured_learning/examples/preprocess)
+subdirectory contains various data preprocessing scripts.

neural_structured_learning/examples/preprocess/cora/prep_data.sh

@@ -13,6 +13,8 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
+# Script to download and preprocess the Cora dataset for use by NSL models.
+
 # URL for downloading Cora dataset.
 URL=https://linqs-data.soe.ucsc.edu/public/lbc/cora.tgz