Analysis of Machine Learning based Feature Selection Method for Sentiment Analysis

An outfit method to include decrease techniques relevant to the production and conduct of research based on the findings would have an important purpose in this article. An outfit strategy means that it is possible to combine at least two strategies. The element reducing approach used is Principal Component Analysis (PCA) for extraction and Pearson Chi-squared factual research findings. The concept investigation is a field where the concepts expressed are understood and organized into positive, negative, and impartial polarities. The highlight is the pivotal machine learning process. This article investigated the implementation, by Naïve Bayes, k-Nearest Neighbor, Support Vector Machine, Logistic Regression, and Random Forest with various Unigram, Bigram, Chi-Square, and Gini Index FSMs, of five machine learning arrangement calculations. However, there has been very little attempt to highlight methods of estimating the Turkish audits. Further component choice strategy, and Question Expansion Ranking, is provided, which will depend on the extended-term weighting strategies used in the field of information recovery to determine mainly important conditions for growth in study.