DESIGN OF WORD-BASED NEGATIVE CONTENT RATING SYSTEM USING NATURAL LANGUAGE PROCESSING

• Main Author: Fathinuddin, Muhammad
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/39795
• Institution: Institut Teknologi Bandung

Rapid development of information technology and internet, information spreads fast and widely. But with the huge amount of information available, it will be difficult to distinguish between positive and negative information. Negative information will create an information distortion when someone reading articles on the internet. Information distortion then have an effect of polarization of meaning which the people won’t succeed in receiving the true meaning of that information. The purpose of this research is to design a text-based negative content rating system that can identify various types of news content from an article on the internet. Articles from internet will be an input and analyzed at word level in the system. Method used in this research is semantic oreientation calculator.for weighting each of words. Usage of this method is to find the meaning of words have to accurately comparing word by words from various type of word database. Sentiment analysis is accurately method for this research purpose. Sentiment analysis can find the type of sentences, is it positive, negative, or neutral sentences. Hate Speech parameter will detect whether the sentence have some strong and very sensitive or inappropriate words that align with government’s policy. While Negative words counter parameter is a comparison of words with the rude words database that contain a negative score. The system will be developed in python environment with the help of Natural Language feature, tokenizer and stemming. Indonesian language are the main language to analyze by the system. Words of Indonesian language are from Indonesia Dictionary with the use of basic words to be used in database. Output of this research is to create a system to analyze what the text in the article gave meaning in each sentences whether contains negative meaning, hate speech, and how many negative words in news article text.