Today, technology is deeply integrated into everyone's lives. Nearly every aspect of modern life, from phone calls to satellites sent into space, has evolved exponentially as a result of technology (Patel et al. 2020a, b, c; Panchiwala and Shah 2020). The advancement of technology has been shaped by the increasing ability to create and manage information. On average, 1826 petabytes are handled over the Internet per day, according to the National Security Agency of the United States (Hariri et al. 2019; Jaseena and David 2014). In light of the rapid increase in data and information communicated over the Internet, it has become necessary to regulate and facilitate the flow of such information (Ahir et al. 2020; Gandhi et al. 2020). These purposes have been addressed by a variety of commercial and social applications. Organizations, governments, and the public can benefit greatly from data and information aspects, including security, research, and sentiment analysis (Jani et al. 2019; Jha et al. 2019). Various optimized techniques are available that are useful for tasks such as classification, summarization, and ease of access and management of data, among others (Shah et al. 2020a, b; Talaviya et al. 2020). Information can be processed using algorithms related to machine learning and deep learning (DL) (Kakkad et al. 2019; Kundalia et al. 2020). Despite the vast amount of information available, computational techniques can allow us to process and analyze information from top to bottom and analyse entire documents or individual words (Pandya et al. 2019; Parekh et al. 2020).
In recent years, the amount of 'natural' data generated by humans has increased dramatically (Shah et al. 2020a, b). The resulting increase in unstructured data has prompted an interest in methods and tools for extracting valuable information from it automatically (Jaseena and David 2014; David and Balakrishnan 2011). In addition to data mining, machine learning, and computational linguistics, text mining is a critical method. Textual analysis can be used to extract information and patterns (Talib et al. 2016b; Fan et al. 2006). It is trivial to do text mining manually, whereby a human reads the text and searches for relevant information. Rather than relying on manual mining, an automatic approach is more logical, as it mines texts at a speed and cost that is efficient (Herranz et al. 2018; Sukhadia et al. 2020; Pathan et al. 2020).
The finance industry is experiencing an increase in unstructured textual data (Lewis and Young 2019). In this context, text mining has a great deal of potential. Kumar and Ravi (2016) examined a number of applications in the financial domain for which text mining can be effective. Their conclusion was that it could be applied to a variety of problems in this industry, including predicting events, managing customer relationships, and maintaining cybersecurity, among others. Over the past few years, many novel methods have been proposed for analyzing financial results. Artificial intelligence has enabled the analysis and even prediction of financial outcomes.
It is noteworthy that finance has always been associated with data, such as transactions, accounts, prices, and reports, since the earliest civilisations. From barter systems to cryptocurrencies, finance has always been associated with data, such as transactions, accounts, prices, and reports. Research and practitioners have increasingly preferred digitized and automated approaches to data analysis and study in the past few decades. Manual approaches to data processing have diminished in importance and relevance. There is significant latent information contained in financial data. The extraction of latent information from a large corpus of data may take many years if it were done manually. As a result of advances in text mining, it is now possible to examine finance related textual data efficiently. Bach et al. (2019) published a review of the literature on text mining for big data analysis in finance. The review was organized into three key areas. These questions referred to the intellectual core of finance, the text mining techniques used in finance, and the data sources associated with the financial sector. Vu et al. (2012) presented a model that utilized text mining on Twitter messages as a method of predicting stock price using sentiment analysis. Kumar and Ravi (2016) discussed this model. They also mentioned the model of Lavrenko et al. (2000), which could classify news stories in a way that could help identify which of them affected the financial markets and to what degree.