Fix broken links to the Neural Graph Learning paper in all NSL documentation.

PiperOrigin-RevId: 299199588

Author: arjung

README.md

@@ -106,7 +106,7 @@ Please see the [release notes](RELEASE.md) for detailed version updates.
 ## References
 
 [[1] T. Bui, S. Ravi and V. Ramavajjala. "Neural Graph Learning: Training Neural
-Networks Using Graphs." WSDM 2018](https://ai.google/research/pubs/pub46568.pdf)
+Networks Using Graphs." WSDM 2018](https://research.google/pubs/pub46568.pdf)
 
 [[2] T. Kipf and M. Welling. "Semi-supervised classification with graph
 convolutional networks." ICLR 2017](https://arxiv.org/pdf/1609.02907.pdf)

g3doc/_index.yaml

@@ -23,7 +23,7 @@ landing_page:
         generated by adding adversarial perturbation have been shown to be
         <b>robust against malicious attacks</b>, which are designed to mislead a model's
         prediction or classification.</p>
-        <p>NSL generalizes to <a href="https://ai.google/research/pubs/pub46568.pdf">Neural Graph Learning</a> as well as <a href="https://arxiv.org/pdf/1412.6572.pdf">Adversarial Learning</a>. The NSL framework in TensorFlow provides the following easy-to-use
+        <p>NSL generalizes to <a href="https://research.google/pubs/pub46568.pdf">Neural Graph Learning</a> as well as <a href="https://arxiv.org/pdf/1412.6572.pdf">Adversarial Learning</a>. The NSL framework in TensorFlow provides the following easy-to-use
         APIs and tools for developers to train models with structured signals:</p>
         <ul style="padding-left: 20px;">
           <li><b>Keras APIs</b> to enable training with graphs (explicit structure) and adversarial perturbations (implicit structure).</li>

g3doc/framework.md

@@ -2,14 +2,13 @@
 
 Neural Structured Learning (NSL) focuses on training deep neural networks by
 leveraging structured signals (when available) along with feature inputs. As
-introduced by
-[Bui et al. (WSDM'18)](https://ai.google/research/pubs/pub46568.pdf), these
-structured signals are used to regularize the training of a neural network,
-forcing the model to learn accurate predictions (by minimizing supervised loss),
-while at the same time maintaining the input structural similarity (by
-minimizing the neighbor loss, see the figure below). This technique is generic
-and can be applied on arbitrary neural architectures (such as Feed-forward NNs,
-Convolutional NNs and Recurrent NNs).
+introduced by [Bui et al. (WSDM'18)](https://research.google/pubs/pub46568.pdf),
+these structured signals are used to regularize the training of a neural
+network, forcing the model to learn accurate predictions (by minimizing
+supervised loss), while at the same time maintaining the input structural
+similarity (by minimizing the neighbor loss, see the figure below). This
+technique is generic and can be applied on arbitrary neural architectures (such
+as Feed-forward NNs, Convolutional NNs and Recurrent NNs).
 
 ![NSL Concept](images/nlink_figure.png)
 
@@ -53,7 +52,7 @@ NSL brings the following advantages:
     shown to outperform many existing methods (that rely on training with
     features only) on a wide range of tasks, such as document classification and
     semantic intent classification
-    ([Bui et al., WSDM'18](https://ai.google/research/pubs/pub46568.pdf) &
+    ([Bui et al., WSDM'18](https://research.google/pubs/pub46568.pdf) &
     [Kipf et al., ICLR'17](https://arxiv.org/pdf/1609.02907.pdf)).
 *   **Robustness**: models trained with adversarial examples have been shown to
     be robust against adversarial perturbations designed for misleading a
@@ -71,7 +70,7 @@ NSL brings the following advantages:
     hidden representations for the "neighboring samples" that may or may not
     have labels. This technique has shown great promise for improving model
     accuracy when the amount of labeled data is relatively small
-    ([Bui et al., WSDM'18](https://ai.google/research/pubs/pub46568.pdf) &
+    ([Bui et al., WSDM'18](https://research.google/pubs/pub46568.pdf) &
     [Miyato et al., ICLR'16](https://arxiv.org/pdf/1704.03976.pdf)).
 
 ## Step-by-step Tutorials

g3doc/tutorials/graph_keras_mlp_cora.ipynb

@@ -75,7 +75,7 @@
         "\n",
         "Graph regularization is a specific technique under the broader paradigm of\n",
         "Neural Graph Learning\n",
-        "([Bui et al., 2018](https://ai.google/research/pubs/pub46568.pdf)). The core\n",
+        "([Bui et al., 2018](https://research.google/pubs/pub46568.pdf)). The core\n",
         "idea is to train neural network models with a graph-regularized objective,\n",
         "harnessing both labeled and unlabeled data.\n",
         "\n",

research/gam/README.md

@@ -53,7 +53,7 @@ Tomkins, S. Ravi. "Graph Agreement Models for Semi-Supervised Learning." NeurIPS
 2019](https://nips.cc/Conferences/2019/Schedule?showEvent=13925)
 
 [[2] T. Bui, S. Ravi and V. Ramavajjala. "Neural Graph Learning: Training Neural
-Networks Using Graphs." WSDM 2018](https://ai.google/research/pubs/pub46568.pdf)
+Networks Using Graphs." WSDM 2018](https://research.google/pubs/pub46568.pdf)
 
 [[3] T. Kipf and M. Welling. "Semi-supervised classification with graph
 convolutional networks." ICLR 2017](https://arxiv.org/pdf/1609.02907.pdf)