|
Title: VoC Classify by SandSIV is an automated text classification engine, one that ach Post by: RealWire on October 16, 2014, 08:52:55 PM VoC Classify by SandSIV is an automated text classification engine, one that achieves a 94% accuracy rate in a live environment.
VoC Classify is a statistical text classification application that has been developed by SandSIV, to provide a high-performance platform for classifying large volumes of customer feedbacks in real-time. In a live environment, with VoC Classify being used by upc cablecom, an affiliate of Liberty Global PLC, it has achieved an accuracy rate of 94%. This represents a significant level of accuracy over and above that which can be achieved by humans or other available tools. Instead of manually classifying documents or hand-crafting automatic classification rules, SandSIV's real-time text categorization uses machine learning methods to learn automatic classification criteria. VoC Classify is a language agnostic application that can be easily trained in an iterative fashion to continue raising the level of accuracy as an ongoing exercise. Speaking on behalf of upc cablecom, Director of Customer Experience Management, David Patarica, has said that, “We were hoping to hit somewhere around 90% accuracy with VoC Classify after some months of training. We actually reached 94% right in the first week – a very impressive result.” Dr. Jukka Hekanaho, CXO and Head of SandSIV Consulting, has stated that, “We have been working on raising the potential accuracy of VoC Classify for some time. To see our clients hitting 94% accuracy is a great tribute to the developers who have designed our statistical classification algorithm, and underlines our ongoing commitment to product excellence.” VoC Classify can be easily activated within SandSIV's Voice of the Customer Hub (VoC Hub™) for existing customers or via an online subscription as a stand-alone service. |