Comparative Analysis of Urdu Parts Of Speech Taggers using Machine Learning Techniques

In modern age technology-based solution can help to maximize consistency of process in bounded discretion-sentencing regimes. Part of Speech Tagging is an exceptionally natural in applications. For example, speech processing, text mining, data extraction and data mining techniques. The language handling of Urdu is a difficult task due to its morphosynthetic uncertainty. The individuals portray, the improvement of a novel tagging approach, utilizing a Conditional Random Field (CRF). In our work we have introduced the Conditional Random Field (CRFs) approach for Urdu Parts of Speech Tagging Model shows a valid accuracy and adjusted language free as a reliant list of capabilities. Our methodology was compared against Recurrent Neural Network Techniques, KNN, Support Vector Machine, Naive Bayes and CART to show that Conditional Random Fields (CRFs) has better accuracy than the other machine learning classifiers.