Understanding language is the most difficult task we undertake as humans and also one of the things that differentiates us from other life forms here on Earth (no letters please about whales, dolphins, chimpanzees or your dog). Computer programs are very specific and recognize only language that has been written into them, so that interacting with them in forms such as a Web interface, a form or button is looking for specific terms or actions. Thus many programs that are made to read unstructured text such as e-mail or Web chat dialogs depend on finding structured text they do understand. Working to overcome the problems that come with unstructured text (the way humans actually think and thus type) is Banter, which has developed its Relationship Modeling Engine (RME), now in version 5.0.
RME is designed to be embedded into software applications to help them understand the communications of customer interactions, to provide further automation and improved routing of critical information. According to Yori Nelken, founder, chairman and CTO at Banter, RME can be used to 'bridge the gap between technical or business jargon and the way customers express themselves,' and improve Web-based self-help solutions. RME is designed to identify the context of a message and identify the intent of the message. It also has embedded learning technology so that users don't have to configure rules as the engine learns from agents and historical data.
When an e-mail or Web form enters the Banter Reply system, Reply's workflow analyzes the message data, identifies the intent of the message and takes the appropriate action. This workflow includes sending the message through RME to categorize what is being asked, and then routing the message to the most appropriate queue. Reply's workflow also allows for such features as sending e-mail acknowledgments when a message is received, auto-formatting and personalization of responses, spell checking, escalation, de-escalation, auditing, tagging of messages that require special assistance, and the tracking of service level agreements. Additionally, incoming messages are assigned a ticket number for tracking and history searching. Banter Reply comes standard with a reporting tool, The Report Center, which can be used for customer trend analysis and agent productivity analysis.
Banter Self Help is designed to provide answers for Web self-service applications. It comes with a reporting feature that allows an administrator to review the questions that are being asked and the answers that are being given. When Self Help is combined with Banter Reply, RME provides for a single knowledge base for both channels, and when feedback is given in one channel, it automatically updates the other channel in real-time.
[ Return To The
Top ]
[ Return
To April 2002 Table Of Contents ]
|