Natural Language Processing Practical using Transformers with Python

Learn how you can perform named entity recognition using HuggingFace Transformers and spaCy libraries in Python. Named Entity Recognition (NER) is a typical natural language processing (NLP) task that automatically identifies and recognizes predefined entities in a given text. Entities like person names, organizations, dates and times, and locations are valuable information to extract from unstructured and unlabeled raw text. At the end of this tutorial, you will be able to perform named entity recognition on any given English text with HuggingFace Transformers and SpaCy in Python. SpaCy is an open-source library in Python for advanced Natural Language Processing (NLP). It is built on the latest research and designed to be used in real-world products. We'll be using two NER models on SpaCy, namely the regular en\_core\_web\_sm and the transformer en\_core\_web\_trf. We'll also use spaCy's NER amazing visualizer. To get started, let's install the required libraries for this tutorial. Fake News Detection in Python: Exploring the fake news dataset, performing data analysis such as word clouds and ngrams, and fine-tuning BERT transformer to build a fake news detector in Python using transformers library. Fake news is the intentional broadcasting of false or misleading claims as news, where the statements are purposely deceitful. Newspapers, tabloids, and magazines have been supplanted by digital news platforms, blogs, social media feeds, and a plethora of mobile news applications. News organizations benefitted from the increased use of social media and mobile platforms by providing subscribers with up-to-the-minute information. Consumers now have instant access to the latest news. These digital media platforms have increased in prominence due to their easy connectedness to the rest of the world and allow users to discuss and share ideas and debate topics such as democracy, education, health, research, and history. Fake news items on digital platforms are getting more popular and are used for profit, such as political and financial gain. It is vital to recognize and differentiate between false and accurate news. One method is to have an expert decide, and fact checks every piece of information, but this takes time and needs expertise that cannot be shared. Secondly, we can use Machine Learning (ML) and Artificial Intelligence (AI) tools to automate the identification of fake news. Online news information includes various unstructured format data (such as documents, videos, and audio), but we will concentrate on text format news here. With the progress of Machine Learning and Natural Language Processing, we can now recognize the misleading and false character of an article or statement. Several studies and experiments are being conducted to detect fake news across all mediums. Paraphrase Text using Transformers in Python: Explore different pre-trained transformer models in transformers library to paraphrase sentences in Python. Paraphrasing is the process of coming up with someone else's ideas in your own words. To paraphrase a text, you have to rewrite it without changing its meaning. In this tutorial, we will explore different pre-trained transformer models for automatically paraphrasing text using the Huggingface transformers library in Python. It includes topics: 1. Named Entity Recognition 2. Fake News Detection in Python 3. Paraphrase Text using Transformers in Python 4. Text Generation 5. Speech Recognition 6. Machine Translation 7. Train BERT from Scratch 8. Conversational AI Chatbot 9. Fine Tune BERT 10. Perform Text Summarization 11. Sentiment Analysis 12. Translate Languages 13. Perform Text Classification 14. Build a Text Generator 15. Build a Spam Classifier I have explained every topic in the most simplest way and you can use these topics in multiple place. Who this book is for: This book is highly appealing to all tech-savvy students, programming enthusiasts, IT graduates, and computer science professionals who want to build strong proficiency in building Python applications. Prior understanding of Python basic coding concepts like variables, expressions, and control structures is required to begin with this book. You can also read Basic Core Python. About the Authors Table of Contents Natural Language Processing Practical using Transformers with Python CHAPTER 1: Named Entity Recognition using Transformers and Spacy in Python NER with Transformers NER with SpaCy Conclusion SourceCode: CHAPTER 2: Fake News Detection in Python Introduction How Big is this Problem? The Solution Data Exploration Distribution of Classes Data Cleaning for Analysis Explorative Data Analysis Single-word Cloud Most Frequent Bigram (Two-word Combination) Most Frequent Trigram (Three-word combination) Building a Classifier by Fine-tuning BERT Data Preparation Tokenizing the Dataset Loading and Fine-tuning the Model Model Evaluation Appendix: Creating a Submission File for Kaggle Conclusion SourceCode: CHAPTER 3: Paraphrase Text using Transformers in Python Pegasus Transformer T5 Transformer Parrot Paraphraser Conclusion SourceCode: CHAPTER 4: Text Generation with Transformers in Python Conclusion SourceCode: CHAPTER 5: Speech Recognition using Transformers in Python Getting Started Preparing the Audio File Performing Inference Wrapping up the Code Conclusion SourceCode: CHAPTER 6: Machine Translation using Transformers in Python Using Pipeline API Manually Loading the Model Conclusion SourceCode: CHAPTER 7: Train BERT from Scratch using Transformers in Python Picking a Dataset Training the Tokenizer Tokenizing the Dataset Loading the Model Training Using the Model Conclusion SourceCode: CHAPTER 8: Conversational AI Chatbot with Transformers in Python Generating Responses with Greedy Search Generating Responses with Beam Search Generating Responses with Sampling Nucleus Sampling Conclusion SourceCode: CHAPTER 9: Fine Tune BERT for Text Classification using Transformers in Python Loading the Dataset Training the Model Performing Inference Conclusion SourceCode: CHAPTER 10: Perform Text Summarization using Transformers in Python Using pipeline API Using T5 Model Conclusion SourceCode: CHAPTER 11: Sentiment Analysis using VADER in Python Conclusion SourceCode: CHAPTER 12: Translate Languages in Python Translating Text Translating List of Phrases Language Detection Supported Languages Conclusion SourceCode: CHAPTER 13: Perform Text Classification in Python using Tensorflow 2 and Keras Data Preparation Building the Model Training the Model Testing the Model Hyperparameter Tuning Integrating Custom Datasets SourceCode: CHAPTER 14: Build a Text Generator using TensorFlow 2 and Keras in Python Getting Started Preparing the Dataset Building the Model Training the Model Generating New Text Conclusion SourceCode: CHAPTER 15: Build a Spam Classifier using Keras and TensorFlow in Python 1. Installing and Importing Dependencies 2. Loading the Dataset 3. Preparing the Dataset 4. Building the Model 5. Training the Model 6. Evaluating the Model SourceCode: Summary