ABSTRACT
When we look at the chapter title, we think for a while, ‘graph’ and ‘language’; whether these terms related to each other? A graph is a pictorial representation, whereas Language is a verbal notion most of the time. The answer is – Yes! Though there are different forms to represent them, there is a strong relationship between these terminologies. A graph is a nonlinear data structure. It is an influential, formal tool to represent and apply multiple aspects associated with language processing. The field of natural language processing or computational linguistics builds on techniques and insights from several different disciplines, principally theoretical linguistics and computer science but with some input from mathematical logic and psychology.
In simple language, it is computing with language. As it looks simple, it is quite hard to implement and map the language by computational skills and techniques. Though it seems to be difficult, it is possible. A graph is not only associated with natural language processing (NLP) but is also used to solve challenges/problems encountered by NLP. In this chapter, we try to provide an overview of how NLP difficulties have been proposed in the graph context, focusing in particular on graph construction – a critical step in exhibiting the data to highlight the targeted phenomena.
A process of mapping some complex object, a textual document, colored image, or a graph into something simple, a fixed length vector e.g., a cluster of numbers or a matrix that captures the significant key features of that complicated object while making it low dimensionalwith fewer objects is called as an embedding.
In this chapter we emphasize graph embeddings and graph embeddings techniques for NLP. This chapter will conclude with a few examples of graph embeddings. This effort may lead to help researchers to solve the problems related to question answering systems, ranking for academic search, well-structured text analysis and classification, fact-checking, explanation regeneration, and many more NLP applications.
