Technology
Latent Metonymical Analysis and Indexing (LMai)

Technology Description

This technology is a self learning algorithm in the realm of Artificial Intelligence that extracts the context or relationships between words automatically (without any guidance given to the machine). LMai is a novel concept of Advanced (Unsupervised) Machine Learning. It uses a mathematical approach to identify the relationships between words in a set of given documents (Unstructured Data). The algorithm examines the natural documents provided to it, automatically makes decisions on clustering related words together and performs the classification. The algorithm does not require any prior training, backed ontology's, or any human intervention during the entire process. This significantly helps in data mining and in building the semantic web.

Benefits

  • LMai supports incremental indexing of data and is not based on SVD.
  • LMai uses a Topical approach during decomposition which is more consistent and accurate.
  • It has shown accuracy/consistency of more than 90% after processing more than 2 Billion Words.

Contact details:

Information Manager
Email: timeis@ficci.com


This Website is Best Viewed in Internet Explorer / Netscape Navigator 1024 X 800 Resolution
Copyright © 2005 Department of Science and Technology <<Disclaimer>>