Analysis system to evaluate and categorize messages/chat content real time

Project Description

The system provides intellectual analysis of incoming texts, conversations and messages to find desired text content defined by experts. Experts in this system create resources for analyzing processes using the standalone GUI application. Experts define desired content to find: words/phrases, expert values, categories and languages. Analyzing part of the system is a cross-platform component for use in high-load processing systems in real-time. It parses and analyzes incoming texts and returns found content which was defined by experts. The Component has a set of additional input parameters, such as custom words/phrases and sensitive level. It returns a list of found text fragments with detailed information: language, probability, category and set of terms. 

Technical features:

• Target platforms: Windows, Linux, Mac OS
• Data sources: Binary storage
• Programming Technologies: Cross-platform C++, STL
• Wrappers: Managed C++ for.Net, JNI for Java

Business features:

• Searching for the desired text content defined by experts within the input texts. 
• High performance of the parsing and analyzing process. 
• Returns found text fragments with detailed information: language, probability, category and set of terms.

Management and resources:

• Timeline (versions 1.0 – 1.3): September 2009 – August 2010 
• Resources: team lead, analysts, software developers and testers
• Development methodology: Agile

See similar projects on our blog.