Bio-Inspired Feature Selection Techniques for Sentiment Analysis – Review

In the recent times, data mining has been the most sought-after area of research. Coming to the text classification domain, lot of research is going into the sentiment analysis, wherein numerous approaches and strategies are used to obtain information about the customer's sentiments. In the past, only supervised, lexicon-based approaches and unsupervised machine learning-based techniques were used to process data. But nowadays, there is a rise in the application of nature-based algorithms. These algorithms are inspired by various social animals who use novel techniques to hunt down their preys or collect food. The very same approach is being simulated into a numerical optimization technique. To understand the same, this paper presents a systematic review of three different bio-mimicking procedures that are currently employed particularly in the arena of sentiment analysis namely Firefly Algorithm (FFA), Cuckoo Search Algorithm (CSA) and Chicken Swarm Optimization Algorithm (CSOA) from the years 2015-2022.