Scalability Issues of NER using Multi-Class Support Vector Machines

Named Entity Recognition (NER) is designed to extract and to categorize rigid designators in written text such as proper names, scientific species, and temporary expressions. There has been increasing interest in this area of research since the early 90's. In this book, we present a pattern shifting away from handcrafted rules, and towards machine learning techniques. Still, latest machine learning techniques have a problem with annotated data accessibility, which is a serious drawback in building and keeping large-scale Named Entity Recognition systems. In this book, we present a new model called as Multi class Support Vector Machine for workflow scheduling in cloud. This workflow scheduling provides a framework for scheduling the entity identification with multiclass Support Vector Machine classifier. The algorithm for the scheduling of resources in cloud called as improved allocation, which continuously and vigorously reallocates multiple types of named entities to the cloud resources to fulfill the cost and performance requirements. This book shows how to create a Multi Class SVM classifier for NER system in environment of cloud.