Table 1.
Common glossaries in natural language processing (NLP).
| Term | Definition | Term | Definition | Term | Definition |
|---|---|---|---|---|---|
| Machine learning (ML) | A branch of computer science dealing with the simulation of intelligent behavior in computers | Deep learning (DL) | An advanced subset of ML that uses layered neural networks to analyze various factors of data | Convolutional neural networks (CNNs) | A type of deep neural network, primarily used to analyze visual imagery |
| Natural language processing (NLP) | A branch of AI focused on enabling computers to understand, interpret, and manipulate human language | Transfer learning | A method in ML where a model developed for a task is reused as the starting point for a model on a second task | Reinforcement learning | An area of ML concerned with how agents ought to take actions in an environment to maximize cumulative reward |
| Vision-language models | Models that understand and generate content combining visual and textual data | Bag of words (BoW) | A simple text representation model in NLP. It treats text as a collection of words regardless of their order or grammar | Corpus | A large collection of text documents or spoken language data used for training and testing NLP models |
| Embedding | Vector representations of words in a continuous space, capturing semantic relationships for NLP model improvement | Feature engineering | Selecting and transforming linguistic features from raw text for NLP model input preparation | Hidden markov model (HMM) | A statistical model in NLP representing sequence data, useful in tasks like speech recognition |
| Information retrieval | The process in NLP of retrieving relevant information from large text collections based on user queries | Jaccard similarity | A measure for comparing the similarity of two sets of words or documents based on shared elements | Keyword extraction | Automatically identifying and extracting key words or phrases from a document for summarization |
| Lemmatization | Reducing words to their base or root form in various word forms like singular/plural or verb tenses | Machine translation | An NLP task for automatically translating text from one language to another computationally | Named entity recognition (NER) | Identifying and classifying named entities (people, places, organizations) in text |
| Ontology | Formally representing knowledge by defining concepts and entities and their relationships in NLP | Question answering | An NLP task of generating accurate answers to questions posed in natural language | Recurrence | Using recurrent neural networks (RNNs) in NLP for processing data sequences in language modeling |
| Sentiment analysis | Determining the emotional tone or sentiment of text, typically as positive, negative, or neutral | Tokenization | Breaking text data into individual units (words, n-grams) for analysis in NLP tasks | Unsupervised learning | Training NLP models on data without explicit labels, allowing independent pattern learning |
| Vector space model (VSM) | A mathematical model transforming text into numerical vectors for similarity calculations in NLP | Word sense disambiguation (WSD) | Identifying the correct meaning of a word in context, especially for words with multiple meanings | Zero-shot learning | Training models to perform tasks they have not explicitly been trained on, used in various NLP applications |