Word Embeddings

Zero-shot classification is the prediction of text content with respect to a given label; the difference in cosine similarity between the word embeddings of words in the text and the label are used for prediction. Word/sentence classification can be performed using several methods: term-frequency, tokenization, embedding. For term-frequency you create…

E-commerce is a boom in the modern business. It is a paradigm shift influencing both marketers and the customers. Rather e-commerce is more than just another way to boost the existing business practices. It is leading a complete change in traditional way of doing business. This significant change in business…

Word embedding is a technique used in natural language processing (NLP) to map words or phrases from a vocabulary to a continuous vector space. The goal of word embedding is to capture the meaning of a word in a vector representation, such that similar words have similar representations and dissimilar…

Paper link: https://arxiv.org/pdf/1907.12009.pdf Conference: ICLR 2019 Contents Summary Weight tying Representation Degeneration Problem Understanding the problem Method to solve the problem Experiment Result My thoughts 1. Summary Representation Degeneration Problem is a phenomena well known to be appearing in a generation task in NLP. When the degeneration happens, word embeddings are not spread…

In this case you are landing here directly, it’s recommended to read through this blog first. In this blog, we shall be using the concept of word embeddings, in order to perform the Sentiment Analysis. We will train our own word embeddings using a simple Keras model for a sentiment…

In recent years, embedding technology to express word vectors has been widely used. Still, there was a problem that words could only be expressed with “points”, and they could not express the spread of concepts and hierarchical relationships. Therefore, an embedded method that expresses data by areas such as “box”…

In Man is to Computer Programmer as Woman is to Homemaker? Debiasing Word Embeddings, which was presented at NeurIPS in 2016 [1], the authors generated embeddings by using Word2Vec trained on a corpus of Google News texts. They then showed how these generated word-embeddings contain biases in their geometry that reflect…