Event classification from the Urdu language text on social media

Area of Science:

• Natural Language Processing
• Machine Learning
• Computational Linguistics

Background:

• The internet's real-time availability has led to a surge in multilingual data on social media and news platforms.
• Effective communication in regional languages presents challenges for event extraction and classification due to resource limitations.

Purpose of the Study:

• To develop and evaluate a machine learning-based event classification system for Urdu text from social media and news channels.
• To address the bottleneck in processing multilingual data for event extraction.

Main Methods:

• Utilized a dataset of over 100,000 labeled Urdu text instances across twelve event types.
• Employed machine learning classifiers, with Term Frequency-Inverse Document Frequency (tf-idf) as the feature vector.
• Features used include the title, its length, and the last four words of a sentence.

Main Results:

• Evaluated six popular machine learning classifiers using tf-idf.
• Random Forest (RF) achieved 98.00% accuracy, and K-Nearest Neighbor (KNN) achieved 99.00% accuracy.
• Demonstrated the effectiveness of the selected features for Urdu event classification.

Conclusions:

• The proposed method offers a high-accuracy solution for event classification in Urdu online text.
• The novel application of specific features (title, length, last four words) enhances event extraction for the Urdu language.
• This research contributes to overcoming resource scarcity in multilingual natural language processing tasks.