Machine learning for automated tender classification
Sumit Goswami, Sunaina Kapoor, Prakriti Bhardwaj · 2011
This paper presents classification of DRDO tender documents into predefined categories. Since there is a consistent growth in the volume of digital documents, both on the internet and within organizations, the need to classify them into categories is obvious. In this paper we used `bag-of-words' technique to represent the tender documents. The dataset was prepared and fed into Weka toolkit. Classification was implemented by Naïve Bayes classifier using 10-folds cross validation technique. The machine resulted in classifying the tender documents with an accuracy of 77.36% by technology category and 67.2% by lab's name.