Pendeteksian Ambiguitas Makna Kata Untuk Meningkatkan Performa Analisa Sentimen Dengan Menggunakan Algoritma Similaritas Path Length

The development of communication and information in this period have been supported with computer network distributed throughout the world, known as the internet. Various information obtained can be useful, the information can be utilized in teamwork to support the decision-making process. The information obtained can be used to identify the sentence by utilizing the techniques of sentiment analysis. Sentiment analysis is a field of study that analyzes the opinions, attitudes, emotions, as well as the evaluation of the language it was revealed in the form of writing. Sentiment analysis is carried out using a method of Lexicon Based Analysis, which will utilize SentiWordnet to resolve the issue of sentiment by way of gathering opinions and do calculation value of sentiment. But problems arise in the use of these methods is the appearance of ambiguity. The exact meaning of election misconduct will result in a mistake the meaning of the sentence as a whole. Problems of ambiguity will be resolved using the similaritas algorithm is the Path Length. The results of this study indicate that Path Length method can improve performance in Precision parameters of 0.5, Recall 0.92, and Accuracy 0.61. An understanding of the context of the sentence is very important, thus the existence of an ambiguity must be resolved in order to find the most appropriate meaning.