hey osam,

in unsupervised learning, you do not have a supervisor which tells you what is right and what is wrong i.e. you do not have input data with example output data where you want to learn the input/output relation.

normally, you use unsupervised learning when you want to find general structures in your data. for example major trends, a low-dimensional embedding or clustering.

do you have a particular problem or can you refine your question?

best,

christoph

In NLP, unsupervised methods are useful for learning representations of words, phrases and sentences, without relying on any labeled data. For instance, word representations (embeddings), such as Glove [1] and Word2Vec [2], are learned using unsupervised methods. Similarly, there are unsupervised methods for learning representations of longer text sequences, such as sentences and paragraphs, e.g. paragraph2vec [3] and skip-thought [4].

Unsupervised methods for representation learning are optimized for tasks, such as reconstruction of the input sequence, and the learned models can then be applied to other tasks. For instance, in [5] the model is learned on the task of predicting the next byte (character) in text, and is later used to predict sentiment expressed in text.

[1] Jeffrey Pennington, Richard Socher, and Christopher D. Manning. 2014. GloVe: Global Vectors for Word Representation

[2] Mikolov, Tomas, et al. "Distributed representations of words and phrases and their compositionality." Advances in neural information processing systems. 2013.

[3] Le, Quoc, and Tomas Mikolov. "Distributed representations of sentences and documents." Proceedings of the 31st International Conference on Machine Learning (ICML-14). 2014.

[4] Kiros, Ryan, et al. "Skip-thought vectors." Advances in neural information processing systems. 2015.

[5] Radford, Alec, Rafal Jozefowicz, and Ilya Sutskever. "Learning to generate reviews and discovering sentiment." arXiv preprint arXiv:1704.01444 (2017).

In Bioinformatics, this can have many applications.