The idea of automatic information extraction from text documents comes from the time of first steps in natural language processing (NLP). Understanding the complex nature of natural language utterances is a key component of Artificial Intelligence. The amount of text data on the web is overwhelming and techniques to extract information automatically from this data will help manage this.
There have been various machine learning models introduced in the past to facilitate NLP tools. Recently, deep learning methods have exhibited relatively high performance in achieving certain tasks for NLP. Deep learning methods employ multiple processing layers to learn hierarchical representations of data, and have produced state-of-the-art results in many domains.
The proposed research will explore and utilise the methods of deep-learning to extract semantic information from the natural language text. From the semantic information, the authors mean that the system is able to (i) detect sentences and their structure (ii) identify named entities (iii) find relations and (iv) present information in a way that support human understanding. On this occasion, the authors have chosen visual representation (graph) of data as an out-put since it is relatively easier to evaluate.
This project is self-funded. Details of studentships for which funding is available are selected by a competitive process and are advertised on our jobs website (View Website) as they become available.