Word Cloud for Online Mobile Phone Tweets towards Sentiment Analysis

Abstract

People became more eager to express and share their opinions on web regarding day-to-day activities or global issues. Social media contributed a transparent platform to share views across the world. Recently research communities, academia, public and service industries are working rigorously on sentiment analysis also termed as opinion mining, to extract and analyze public mood and views. Data pre-processing is a crucial step in sentiment analysis and selecting an appropriate pre-processing methods can improve classification accuracy. In this paper, we explore the role text pre-processing of online mobile phone reviews towards Sentiment Analysis. Proposed text pre-processing methods remove inconsistent and redundant elements on the collected data to improve classification accuracy. Proposed Preprocessing methods involves removal of punctuations, special characters, digits, escaping HTML characters, decoding data, Apostrophe Lookup, Removal of Stop-words, Removal of URLs, Removal of Expressions. The final pre-processed online review data are presented in the form of word cloud with the frequency statistics of the keywords.

Analisis Speech-to-Text Pada Video Mengandung Kata Kasar Dan Ujaran Kebencian Dalam Ceramah Agama Islam Menggunakan Interpretasi Audiens Dan Visualisasi Word Cloud

“…Harapannya dapat memantau perilaku pengguna media sosial yang melakukan kejahatan ujaran kebencian terhadap individu tertentu. Penelitian lain pada tahun 2019, tentang Analisis sentimen dari Twitter yang berfokus pada ulasan suatu produk dengan menggunakan word cloud oleh Naramula Venkatesh dan A.Kalaivani [5]. Penelitian ini menggunakan metode term document matrix untuk mengolah data yang telah terkumpul dan divisualisasikan menggunakan word cloud.…”