Classification of Competition Content on Web Pages using a Machine Learning Algorithm

To develop quality and human resources in the future, students as an agent of change can increase their potential by participating in competitions. There are so many benefits that can be obtained from participating in competitions such as increasing abilities, improving soft skills, increasing experience, expanding relationships, and of course, getting prizes and certificates. Competition information can be found on the internet, especially in search engines. As the web continues to evolve, it is becoming increasingly difficult to find relevant information using traditional search engines. But even though it has been optimized by SEO, it is not uncommon to find web pages that provide invalid competition information and pass the deadline. So, a web page classification is needed. This paper uses a machine-learning algorithm to classify valid competition information web pages for students with the Gradient Tree Boosting classifier method. The Gradient Boosting Algorithm is one of the more sophisticated ensemble methods for online boosting classifiers, which often improves classification accuracy. To do machine learning with this algorithm we split the dataset into two parts, which are for testing 30% and 70% for training. And the parameters are learning_ rate=0.01, n _ estimators=150. Then, we got 85% accuracy. This result is good for classification and proves that the machine learning algorithm can be used as a good method for web classification because it is more effective and efficient to retrieve valid competition information rather than manually validating. According to the result, our proposed method shows a promising performance to validate the competition information.