Natural Language Processing - Basic to Advance using Python

Learn basic string operations in Python for natural language processing, including splitting lines, joining strings, and simple text manipulation to prepare and transform text data.

Explore the basics of regular expressions in Python, learning to craft and test patterns for line and character matching, searches, and practical regex skills.

Install and test NLTK for Python using Mac and Unix steps, set up the folder, download dependencies, run the installation script, and verify components are ready for learning.

Explore NLTK tokenizers and how tokenization and sentence segmentation vary by language, addressing punctuation and capitalization challenges while extracting sentences and words from text using practical examples.

Explore part-of-speech tagging with NLTK in Python, tagging words with POS labels and understanding noun, verb, and adjective roles. See practical tagging examples and how grammar underpins sentence structure.

Learn the basics of nltk stemming and lemmatization, including how to convert words to their dictionary forms, compare snowball and lancaster stemmers, and normalize text data.

Learn word sense disambiguation by using context to determine a word's meaning, illustrated with bank as a financial institution or river bank, and how definitions help resolve ambiguity.

Discover how nltk bleu scores evaluate translation quality by comparing a candidate translation to a reference, and compute a score across sentences or paragraphs.

Explore Stanford NLP by downloading and unzipping a package, inspecting the folder structure, and running tokenization and part-of-speech tagging on sample sentences to understand word-level tagging in natural language processing.

Learn how to use TextBlob in Python to perform tokenization, sentence segmentation, part-of-speech tagging, noun phrase extraction, sentiment analysis with polarity and subjectivity, and language detection and translation.

Learn privacy considerations in web data handling, explore web crawling and content download with libraries, and follow legal cautions by practicing only on authorized sites such as Wikipedia.

Load and clean a review dataset for natural language processing, focusing on the top five products. Export a two-column subset containing score and text for modeling.

This lecture demonstrates standard string cleaning for text data, removing noise and extra whitespace, converting text to lowercase, and using regex to replace patterns with spaces.

Learn to clean text data by removing punctuation and other defects, filter tokens using stop words, and apply spelling corrections to prepare English text for NLP models in Python.

Learn practical string cleaning for NLP, covering spelling correction, lemmatization, stemming, and removing noisy tokens like single letters and long words, plus data saving and offline checks.

Learn to build and customize a word cloud from clean text data in Python, covering data cleaning, text preprocessing, and visual styling with font and background colors.

Explore the six-phase cross-industry standard process for data mining to plan NLP projects, including business understanding, data preparation, modeling, evaluation, and deployment, using clean data from the basics.

Understand entity resolution and deduplication to link records across name, address, and attributes, enabling accurate data normalization using similarity measures.

This lecture covers entity resolution and deduplication with a record linkage library, showing data prep, handling missing values, and the trade-off between full versus indexed matching for large datasets.

Learn how blocking and indexing optimize single-table entity resolution, select key comparison features, apply distance metrics, and use unsupervised or supervised methods to identify duplicates.

Learn entity resolution and deduplication across two tables by applying blocking, column-level comparisons, and match decisions to identify and group duplicates in data.

Explore how text is transformed into features using one hot encoding in Python, turning words into indicators for machine learning models.

Explore how count vectorizer builds a vocabulary, counts word occurrences, and creates text features, including single words and two-word phrases.

Learn how tf-idf, combining term frequency and inverse document frequency, weighs terms to filter out common words and highlight important words across documents for effective text mining.

Learn how word embeddings map words to low-dimensional vectors that capture context and semantic similarity, using word2vec and GloVe with cbow and skip-gram variants.

Explore word2vec and GloVe by loading models, inspecting vocabulary, and performing word analogies and similarity checks to understand embedding dimensions and vector relationships.

Explore word embeddings from custom review data using natural language processing in python, covering data preparation, model training, and analysis of word relationships through dimensional representations.

Explore word sense disambiguation in natural language processing by using context to distinguish meanings of ambiguous words like bank, with practical examples in Python.

Explore speech recognition with a microphone using Python libraries and the Google API to convert speech to text, including offline options and hidden Markov model basics.

Practice speech recognition in natural language processing by converting audio files to text in Python. Use language options such as Hindi and address background noise to improve accuracy.

learn how to measure similarity between two strings using cosine similarity and other distance metrics, including cosine distance and edit-based distances, with practical examples.

Learn language translation with Python using the Google API package to translate online or offline, explore supported languages, and work with Hindi and French examples for NLP practice.

Explore how computational linguistics blends grammar, syntax, and statistical modeling to study language. Apply ambiguity resolution, dependency parsing, and named entity recognition in natural language processing tasks like machine translation.

Learn dependency parsing in computational linguistics, turning sentences into token-based dependencies and observing their connections in a local-host display. Practice with NLP libraries to generate and inspect dependency structures.

Learn how machine learning and deep learning apply to natural language processing, with data processing, feature selection, and modeling. Explore applications like sentiment analysis, classification, data extraction, and topic modeling.

Apply random forest classification to real data, train and fit models, tune key parameters such as max depth and feature selection, and interpret accuracy and prediction probabilities for decision making.

Explore classification in machine learning using Naive Bayes and XGBoost, compare predictions, assess accuracy, and tune hyperparameters to improve model performance on real datasets.

Explains how to build a deep learning classification model using tf.keras mlp for a five-class task, and evaluates it with a confusion matrix and around 92-93% accuracy.

Leverages tf.keras's built-in embedding layer to perform multiclass text classification with deep learning, converting sentences into embedded sequences and training the model on cleaned data for prediction.

Learn to build a text classifier using deep learning by averaging word vectors into 100-dimension embeddings with tfkeras, then concatenate features and train a model to evaluate accuracy.

Explore deep learning classifications using tfkeras with a custom word vector, including text preprocessing, lowercasing, numeric encoding, and one hundred word representations.

Explore bias–variance tradeoffs to decide if a model is good enough, distinguish high bias and high variance scenarios, and improve with more data, features, and regularization in business terms.

Explore sentiment analysis in natural language processing by classifying social media feedback into positive, negative, and neutral, using machine learning, deep learning, and rule-based, dictionary-based approaches, with polarity insights.

Explore k-means clustering in natural language processing by comparing distances from points to centers, assigning points to the nearest center, and forming clusters.

Explore topic modeling in natural language processing with Python to uncover underlying topics in large text collections, cluster documents, and extract meaningful topics.

Explore building a semantic search engine using cosine similarity and a dictionary of words to compute relevance and rank results by word distance.

Explore real-world natural language processing applications, from portfolio management analytics to language and vision integration, and study morphology, lexical analysis, ontology, and semantic meaning.