In this node, a discussion about word2vec's two important tricks is provided. Additionally, a explaination about why these two tricks are needed is provided as well.
1. Hierarchical Softmax
As known, word2vec uses the hidden layer gotten by multiplied with the W matrix as the representation vectors of words. For a naive model, we calculate the average of the word vectors as an output. As a result, we need do it for every word and do normalization on it. It's very time consuming.
Hierarchical softmax, instead, provides a Huffman tree based model. And for every node, we define a logistic regression which decides the following direction.(1 or 0 means left or right.) And by considering the path as products of the possibilities, we could calculate the probabilities of every word reaching the leaf node. And we calculate every partial derivatives for every node going through by a path sample. Finally, we update the parameters according to the derivatives. It highly reduces the operand.

2.Negative Sampling
This trick is provided for solving the problem of sampling and dealing with the commonly used words with no sense, like $the$. For every words in the samples, there exists a probability of deleting the relating pairs positively related with the frequency it shows up. And five negative word pairs are randomly provided for every sample for updating. What's interesting? The possibility of a word selected as a negative word is also related with the frequency it shows up. 

Tricks Used in Word2Vec

Word2vec is a technique for natural language processing. The word2vec algorithm uses a neural network model to learn word associations from a large corpus of text. Once trained, such a model can detect synonymous words or suggest additional words for a partial sentence.

University of Michigan - Ann Arbor

1. LSA based method.
This kind of method is based on the intuition of SVD similar way. They try to do SVD on term-document matrices so that get the representation vectors.
2. local context based method.
This kind of method try to use the information gotten from the local context.
e.g.Word2Vec,Glove(Glove is actually a combination of those two methods.)

Algorithms used in word embedding

https://www.coursera.org/learn/nlp-sequence-models

Sequence Models

Word2Vec

GloVe is an unsupervised learning algorithm for obtaining vector representations for words. Training is performed on aggregated global word-word co-occurrence statistics from a corpus, and the resulting representations showcase interesting linear substructures of the word vector space.

Glove

Elmo is a well-known word embedding model as well. It uses CNN to transfer words to raw word vectors. Then it uses bi-lstm and calculate the weighted average vectors of the raw word vectors and the two hidden layers to get the final word vectors. It's proposed as a pre-trained language model mainly to solve polysemy problem.

Elmo

https://en.wikipedia.org/wiki/Latent_semantic_analysis

Latent Semantic Analysis(LSA)-Reference

LSA is a well-known and widely used language model in natural language processing area.
LSA assumes that words that are close in meaning will occur in similar pieces of text (the distributional hypothesis). A matrix containing word counts per document (rows represent unique words and columns represent each document) is constructed from a large piece of text and a mathematical technique called singular value decomposition (SVD) is used to reduce the number of rows while preserving the similarity structure among columns. Documents are then compared by taking the cosine of the angle between the two vectors (or the dot product between the normalizations of the two vectors) formed by any two columns. Values close to 1 represent very similar documents while values close to 0 represent very dissimilar documents.
It's the basic method of Glove and many other modern methods of word embedding.

Latent Semantic Analysis

The key idea of noise constractive estimation is to learn the feature of data distribution by learning from a dichotomy problem of selecting the data points from the original dataset from a mixed dataset made up of noises and the original datasets.

Noise Contractive Estimation

Ranking loss focuses more on relative distance instead of measuring the direct difference between the output with the expectation. 

Ranking Loss

Ensemble learning is a simple and effective technique for building a strong predictive model by combining the outputs of multiple, often weaker, individual models. This method aggregates the predictions from these constituent models to achieve a more robust and superior performance than any single model could attain on its own.

Ensemble Learning

BERT (Bidirectional Encoder Representations from Transformers) stands out as one of the most popular and extensively used pre-trained sequence encoding models in the field of Natural Language Processing. As a foundational model, it is trained using a self-supervised approach that combines two tasks: masked language modeling (MLM) and next sentence prediction (NSP). In MLM, the model predicts randomly masked words from their context, enabling it to learn deep bidirectional language representations. This dual-task training makes BERT a versatile foundation model adaptable to a wide array of NLP applications.

Learn Before

Related

Learn After