User Tools

Site Tools


colloquium:treparel_text_analytics_cluster_classify

Treparel Text Analytics: Cluster, Classify & Visualize

Dr. Anton Heijs (Treparel) (2014-07-08 13.00-13.45 in ZI-2126)

KMX Text Analytics provides ad hoc access to explore through large sets of text and content. It does not require schema’s or pre-structured repots. Instead, KMX extracts the elements of meaning from documents and other unstructured data sources and combines it at query or search time. In short KMX:

  • helps to disambiguate terms by using each term’s context
  • is language-based or meaning-based so it can understand the meaning of a query, document, phrase, paragraph
  • is designed for instant and limited human user interaction (machine learning)
  • supports fuzzy or probabilistic matching to find relevant documents that exact matching miss
  • can work with knowledge bases that include dictionaries for meaning like taxonomies or ontologies or that can help you to extract such knowledge bases by extracting entities from the text corpus
  • supports exploration or sorting a large set of potentially relevant information by the use of relevance ranking, automatic clustering, machine learning based categorization, faceted search, interactive visualizations
  • that provides data discovery through an intuitive interface to help users to explore data without much training
colloquium/treparel_text_analytics_cluster_classify.txt · Last modified: 2014/07/03 11:59 by Mena B. Habib