Nowadays, exponential increase in the social media users on different online social networking platforms has led to the huge number of information exchange and views expressions among millions of users in the form of posts, comments, and chats. This has also led to the need of sentiment analysis where user sentiments are predicted, and corresponding actions are taken using machine and deep learning techniques. The leading social media platform used for opinion mining is Twitter, which contains a huge volume of the text that is used as the most emerging research area for sentiment analysis by mining tweets to capture sentiments. In current times, sentiment analysis on political tweets has observed an explosive growth which considers political tweets, election data, public tweets regarding policies and their political sentiments. In previous research works, neural networks have been used to analyze political sentiments considering text obtained from a large number of twitter handles. The twitter dataset is used to evaluate through different ML and DL techniques for political sentiment analysis by conventional researchers in recent years. In this article, an evaluation of about 36 studies is included, published during the period from the year 2010–2024. The article mainly considered leading machine learning approaches such as SVM, Naïve Bayes, and Random Forest along with deep learning approaches such as LSTM, BERT, and RoBERTa popularly used for political sentiment analysis. ML-based classifiers mainly applied with feature representation methods such as Bag of Words (BoW) and Term Frequency-Inverse Document Frequency (TF-IDF) for sentiment analysis on benchmark political tweet datasets. Ultimate aim of study is to identify appropriate methods for future investigations by comparing the features, limitations, datasets, and result parameters and determine research gaps. Furthermore, our literature review observed that majority of studies have used deep learning methods that enhanced a substantial amount of performance of sentiment analysis on employed datasets.