Dr. Mohammed Ali Alzahrani, International Journal of Computer Science and Mobile Computing, Vol.9 Issue.1, January- 2020, pg. 84-99 © 2020, IJCSMC All Rights Reserved 84 Available Online at www.ijcsmc.com International Journal of Computer Science and Mobile Computing A Monthly Journal of Computer Science and Information Technology ISSN 2320–088X IMPACT FACTOR: 6.199 IJCSMC, Vol. 9, Issue. 1, January 2020, pg.84 – 99 Odd Posts Identification through the Vocabulary by Semantic Sentiment Analysis Using Machine Learning Algorithm Dr. Mohammed Ali Alzahrani College of Computing and Information Technology, Taif University, Saudi Arabia marzahrani@tu.edu.sa Abstract— In this work, the importance to detect the odd posts is considered to understand. This work explains the odd posts, and the methods using machine learning algorithms are well explained. In the first part, a survey is conducted on what runs on the internet and why to eliminate the odd posts is required. After explaining the importance of the odd posts as the data is exponentially increasing, to eliminate the odd posts various techniques are used but here are some of the best, TF-IDF is used as the clustering technique, SVM used to eliminate the odd posts and observed as a better solution. However, to eliminate the odd posts using machine learning algorithms RNN produces good results, as RNN is feasible to use for the large scale dataset. The sentiment analysis using machine learning approaches is used to obtain the progressive results which aim to eliminate the odd posts from the data, as the current demand and major source of the data are social media platforms. The accuracy is improved, and the implementation of the model is done on the dataset of the tweets which includes negative, positive and neutral tweets. Keywords— Odd posts identification, semantic sentiment analysis, machine learning, TF-IDF, Support vector machine, Recurrent neural network. ____________________________________________________ Introduction: The last decade had increased the use of the internet data, and the internet-based posts and especially the textual data increased. The exponential increase in the textual data, the huge issues regarding fair communication, arise in meanwhile machine learning plays its part, by developing such methods to analyze the textual data. The content requires fulfilling the term and remaining ethically strong enough to remain in a good category. However, this work explains to develop the dictionary or the vocabulary to help ease in detecting the odd words first. The current era and