ner spacy readme added

README.md

@@ -504,6 +504,20 @@ The transformer alone is not giving good results compared to BiLSTM in the NER t
 ![ner](assets/images/applications/classification/ner_transformer_crf.png)
 
 
+### Day 89: NER tagging with Spacy
+
+SpaCy provides an exceptionally efficient statistical system for NER in python, which can assign labels to groups of tokens. It provides a default model which can recognize a wide range of named or numerical entities, which include person, organization, language, event etc.
+
+
+![ner](assets/images/applications/classification/ner_spacy.png)
+
+Apart from these default entities, spaCy also gives us the liberty to add arbitrary classes to the NER model, by training the model to update it with newer trained examples.
+
+2 new entities called **`ACTIVITY`** and **`SERVICE`** in a specific domain data (bank) are created and trained with few training samples.
+
+![ner](assets/images/applications/classification/ner_spacy_custom.png)
+
+
 Checkout the code in `applications/classification` folder
 
 ## Generation based Applications

applications/classification/README.md

@@ -179,3 +179,11 @@ After trying NER tagging with Transformer Encoder, NER Tagging with pre-trained
 ## NER tagging with Transformer-CRF.ipynb
 
 The transformer alone is not giving good results compared to BiLSTM in the NER tagging task. Augmenting the CRF layer on top of the transformer is implemented which is improving results compared to standalone Transformer.
+
+## NER tagging with Spacy.ipynb
+
+SpaCy provides an exceptionally efficient statistical system for NER in python, which can assign labels to groups of tokens. It provides a default model which can recognize a wide range of named or numerical entities, which include person, organization, language, event etc.
+
+Apart from these default entities, spaCy also gives us the liberty to add arbitrary classes to the NER model, by training the model to update it with newer trained examples.
+
+2 new entities called **`ACTIVITY`** and **`SERVICE`** in a specific domain data (bank) are created and trained with few training samples.

applications/classification/ner_tagging/README.md

@@ -104,3 +104,28 @@ After trying NER tagging with Transformer Encoder, NER Tagging with pre-trained
 The transformer alone is not giving good results compared to BiLSTM in the NER tagging task. Augmenting the CRF layer on top of the transformer is implemented which is improving results compared to standalone Transformer.
 
 ![ner](../../../assets/images/applications/classification/ner_transformer_crf.png)
+
+
+## NER tagging with Spacy.ipynb
+
+SpaCy provides an exceptionally efficient statistical system for NER in python, which can assign labels to groups of tokens. It provides a default model which can recognize a wide range of named or numerical entities, which include person, organization, language, event etc.
+
+
+![ner](../../../assets/images/applications/classification/ner_spacy.png)
+
+Apart from these default entities, spaCy also gives us the liberty to add arbitrary classes to the NER model, by training the model to update it with newer trained examples.
+
+2 new entities called **`ACTIVITY`** and **`SERVICE`** in a specific domain data (bank) are created and trained with few training samples.
+
+![ner](../../../assets/images/applications/classification/ner_spacy_custom.png)
+
+Many annotation tools are available for creating the training data. Few of them are:
+
+- [prodigy](https://prodi.gy/)
+- [doccano](https://github.com/doccano/doccano)
+- [inception](https://inception-project.github.io/)
+
+#### Resources
+
+- [Spacy documentation on training ner](https://spacy.io/usage/examples#training-ner)
+- [video tutorial](https://www.youtube.com/watch?v=1ePkOSGoIFI&list=WL&index=7&t=0s)