In this post , I am going to restrict the scope of analytics to just text analytics and how this becomes the starting point of information discovery for intelligent systems. This is really about how NLP helps power ingestion for text analytics.
Text Analytics or Text Mining is the discovery of new or existing facts applying the science of natural language processing and statistical learning techniques. Below is the standard text analytics pipeline that is used to understand documents and then create a knowledge repository of textual content that can further be used for purposes to search , index , categorize , observe trends or analyze and represent visually.
The reference architecture of the text analytics systems that are based on Natural Language Processing need to work with three other distinct components
Knowledge Repository – Where the intelligent systems builds factoids, hypothesis – essentially stores and transforms knowledge into processable chunks.
Inference Engine – NLU , Semantics is essentially part of this layer that helps the system reason and predict and uses a variety of AI based technology to help the system extract meaning out of content based on context.
Interaction Layer– The layer which has augmented intelligence built in to collect and self learn in order to infer and plug into interfaces designed to interact with humans.
In most banks today , the focus is primarily the integration to such analytics to channel solutions and the UX of existing apps. However , focus on building the knowledge repository is lesser for banks with the ‘overnight digital’ strategy but some banks have been working dedicatedly over the last 4 years or more to build the knowledge repository and create the right reference models making the inference engine far more evolved . This helps the interactions to be far more intuitive . Context Training and Corpus are in other words the real differentiators for banks. More on the vendors , banks and use cases in a later post but till then if you want to catch up on the basics and understand NLP in greater detail you can read my earlier post – What is Natural Language Processing ?
For further posts on the topic , make sure you follow the blog :).