Cluster Analysis

Overview

Cluster analysis is an unsupervised technique that groups similar data points into "clusters," where items within a cluster have similarities that distinguish them from items in other clusters. Linguistic clustering takes place by calculating distances or similarities between term frequencies, without reference to any pre-existing classification labels. It can be used to identify latent patterns that can be hard to observe by the human eye.

Lexos provides two different ways to cluster your documents. This page is a starting point for exploring the different methods.

• Hierarchical Agglomerative Clustering
• K-Means Clustering