Data Classification Using Machine Learning Approach

摘要

Currently, Internet has numerous effects on our everyday lifecycle. Its significance as an intermediate for commercial transactions will develop exponentially throughout the next years. In terms of the engaged marketplace volume, the Business to Business region will hereby be the supreme exciting area. As the extensive usage of electronic business transactions increase, great volume of products information gets generated and managing such large information automatically becomes a challenging task. The accurate classification of such products to each of the existing classes also becomes an additional multifarious task. The catalog classification is an essential part for operative electronic business applications and classical machine learning problems. This paper presents a supervised Multinomial Na?ve Bayes Classifier machine learning algorithm to classify product listings to anonymous marketplaces. If the existing products are classified under the master taxonomy, the task is to automatically categorize a new product into one of the existing categories. Our algorithm approach proposes a method to accurately classify the existing millions of products