What we’re thinking about

Insights, news, and tips from our top tech and business innovators.

Machine learning is the new search (part 3)

Annelore van der Lint | August 10, 2017

With the ever-growing volumes of data, machine learning methods designed to automate data analysis are indispensable.

Finding what you did not know - topic modelling and clustering

Topic Modelling and Cluster Analysis are two approaches to text-mining. A topic model is used to statistically explore abstract concepts (topics) that occur within a set of documents. Cluster analysis uses perceived relation between various groups of objects to create new sub-groups (clusters). These documents are ideal to submit for Machine Learning.

ZyLAB’s solutions use Topic Modeling to automatically generate an overview of the most used concepts or topics in a text collection. These algorithms automatically find the most dominant topics in a document set and use for each topic the best words to describe them. This works completely un-supervised. By clustering and visualizing the found topics in a hierarchical tree or by using an interactive Word Wheel, you get immediately a clear overview of the dominant topics.

Topic modeling is a useful method that enhances the users’ ability to interpret large volumes of information. With these techniques, you can actually find relevant documents even if you did not really know what words or topics to look for.

A Hierarchical Tree and a Word Wheel Visualization of Automatically Found Topics

Word wheel

All blogs in this "Machine Learning is the New Search" series: PART 1 | PART 2 | PART 3

Written by Annelore van der Lint

Connect with Annelore van der Lint on LinkedIn or