Text Analytics and Pattern Mining

Overview

Text constitutes almost 80% of all data that is generated today and contains valuable insights that influence business outcomes in a significant way. Datamatics’ solutions efficiently mine voluminous text and huge data sets to understand, analyse, and manage the entire spectrum of multi-structured data and free flowing text for taking real-time decisions. This includes spread sheets, documents, PDFs, images, video, audio, and any other streaming or static data from news feeds, emails, social media etc. Datamatics offers two types of analytics solutions – Text Analytics and Pattern Mining.   

What do these solutions do?

LexiconTM is Datamatics’ Text Analytics and Artificial Intelligence product. LexcionTM based solutions make use of Cognitive Algorithms to parse millions of unstructured documents, summarise them, extract key and relevant information, classify them, and perform contextual analysis to derive intelligence. These solutions support multiple languages and help enterprises to unearth actionable insights from Terabytes of free flowing text. Backed by fast analytics engines, the solutions provide output in the form of easy-to-use smart dashboards. 

SynapsisTM is Datamatics’ Pattern Mining product. SynapsisTM based solutions help businesses gain deep insights with high precision as compared to fuzzy logic based solutions. These solutions use Advanced Analytics and highly evolved Genomics that discover deep relationships in large data sets and seemingly diverse data points. Solutions based on this product discover patterns and predict new sequences to unearth opportunities and pre-empt potential threats. The solutions can detect associations and rules and discover current transaction patterns as well. These solutions are driven by high performance analytics engines and support multiple languages. With ingrained self-learning characteristics, the solutions evolve as new patterns get detected. 

Datamatics’ Advantage:

Datamatics’ solutions use sophisticated Evolutionary Algorithms, Natural Language Processing, Advanced Neural Networks, Artificial Intelligence, and Machine Learning Algorithms. With an objective to turn complex data into insights, the solutions automate the challenging task of extracting intelligence from semantics in both static and streaming data. The solutions use high performance analytics engines and support multiple languages. These solutions have a highly interactive visual interface, which can be used to run different queries. Some of the solutions include:

  • Document Classification solutions, which can be used to auto-classify textual data for information classification, routing, text scrutiny, etc.

  • Email Classification solution, which analyzes emails for specific action to be taken based on the content of the email. It integrates with Robotics for any further sequential actions that need to be taken.

  • Text Mining solutions, which can intelligently extract topics, entities, sentiments, and specific information from free text and multi-structured data sources. The solutions can further auto-summarise, associate, and co-relate the named entities.

 

Key Technologies:

  • Seamlessly connects to most of the Big Data Platforms, such as Hadoop HDFS, MapR, MongoDB, Solr, Neo and Titan as well as standard RDBMS, such as MSSQL and Oracle

  • Uses DRILL to query the database


  • Uses an NLP engine to derive insights from text

  • Uses Artificial Intelligence and Machine Learning to understand and extract meaning from text, co-relate text instances, summarise and provide elements for inference drawing and decision making