Book Image

The Handbook of NLP with Gensim

By : Chris Kuo
Book Image

The Handbook of NLP with Gensim

By: Chris Kuo

Overview of this book

Navigating the terrain of NLP research and applying it practically can be a formidable task made easy with The Handbook of NLP with Gensim. This book demystifies NLP and equips you with hands-on strategies spanning healthcare, e-commerce, finance, and more to enable you to leverage Gensim in real-world scenarios. You’ll begin by exploring motives and techniques for extracting text information like bag-of-words, TF-IDF, and word embeddings. This book will then guide you on topic modeling using methods such as Latent Semantic Analysis (LSA) for dimensionality reduction and discovering latent semantic relationships in text data, Latent Dirichlet Allocation (LDA) for probabilistic topic modeling, and Ensemble LDA to enhance topic modeling stability and accuracy. Next, you’ll learn text summarization techniques with Word2Vec and Doc2Vec to build the modeling pipeline and optimize models using hyperparameters. As you get acquainted with practical applications in various industries, this book will inspire you to design innovative projects. Alongside topic modeling, you’ll also explore named entity handling and NER tools, modeling procedures, and tools for effective topic modeling applications. By the end of this book, you’ll have mastered the techniques essential to create applications with Gensim and integrate NLP into your business processes.
Table of Contents (24 chapters)
1
Part 1: NLP Basics
5
Part 2: Latent Semantic Analysis/Latent Semantic Indexing
9
Part 3: Word2Vec and Doc2Vec
12
Part 4: Topic Modeling with Latent Dirichlet Allocation
18
Part 5: Comparison and Applications

The Ensemble LDA for Model Stability

One of the success criteria of topic modeling is to produce a reliable set of topics. However, many experiments with Latent Dirichlet Allocation (LDA) have shown that the topics can be unstable and not reproducible. This issue seriously limits the applications of LDA. The instability of the topic results is partly due to the fact that the model settles at a local maximum depending on the random initialization. Even if a seed number is set to control random initialization, noisy topics can be generated during the modeling process, which might influence the quality of the outcome.

The root cause of the instability is that a single LDA model identifies the “true” topics and “pseudo” topics and produces noisy predictions. If the model is trained again, it will identify “true” topics and other “pseudo” topics. The solution is to build multiple models or an ensemble of models to weed out the pseudo...