Article Preview
Top1. Introduction
Among all the social websites, widely used site is Twitter where the users share and post their opinions according to their interest such as sports, education, transport, research area, medical and health, politics and information related to the share markets through tweets. Share holders make investments and undergo through huge profits if they are knowledgeable about the company's stock values. Otherwise, they have to incur heavy losses and may lose lifelong savings.
India's premier stock exchanges are the Bombay Stock Exchange and the National Stock Exchange (Stock Market, n.d.). Stock market analysis, which is the evaluation of a market as a whole, is done to take a proper decision to incur better profits by investing in a suitable firm (Stock Analysis & Investopedia, n.d.). There are 2 ways in which the analysis can be carried out. The first is a fundamental analysis, where in the country's economical and financial conditions are assessed to make a decision about investment based on the balance sheet, profit and loss statements etc. On the other side, there is technical analysis, which is based on the supply-demand analysis and historic data analysis independent of the financial aspects around. Customer can choose a suitable one based on the knowledge levels acquired, trend analysis and formula to achieve better return on investments (Technical & Fundamental Analysis of Share Market, n.d.).
(Richard, Tobias and Willy, 2018) at the other end showed that, in addition to the focus on trends in stock market, it is also essential to gather inputs on market resiliency, which is the worth of processing a transaction with a minimal impact on the cost factor, in accordance with the elasticity of supply and demand in the market.
Based on the reviews of a particular product like electronic gadgets, wrist watches, wall decorators etc. in the social networking websites, a person may prefer to purchase it. (Ajinkya, Anjali, Anita and Shriya 2015) showed in their work that the aforementioned approach works well for a limited set of products and a limited set of companies forecasting the reviews using available tools and techniques. Positive feedbacks obtained on a particular product will attract huge set of audience to go ahead with the review decisions, there by strengthening the necessity to use the product. At the other end, if the feedback is negative, it enables the designers of the product to re-iterate on the working model and overcome the flaws.
To perform sentiment analysis on twitter data, the relevant API is used that enables the developers to access nearly 1% of tweets at a particular timestamp, based on an appropriate keyword (Mahajan & Rana, 2018). A tweet usually comprise of plain text, emoticons, user name, location and time stamp as retrieved by the twitter API as used by the authors (Anjali & Ajay, 2017).
Biggies around the world like Google, Microsoft, Facebook, Yahoo etc provided a wide prospect to deal with twitter streaming data. Steps are formulated to carry out the analysis on any local machine. The only pre-requirement for a user is to have a twitter account. It is also essential for the user to create a namespace, resource group, event group and obtain the access permissions to the same. Once the required groups are created, a stream analytics job can be created where it is essential to specify the nature of input parameters, query and output sink (Real-time Twitter sentiment analysis in Azure Stream Analytics, n.d.).
Towards this end, the goal of this work is to address the challenge of providing better inputs to the customers interested to invest in the share market to earn better returns on investment. The work is developed using a web crawler to speed up the rate at which data is gathered across a variety of socio-websites generating tweets with huge set of contents including text, emoticons etc. However, the data at this stage is incomplete and may contain missing and noisy data. Hence, pre-processing techniques are applied to filter the data and remove unwanted characters that intend to convert the tweets into a meaningful statement. These tweets if contains quotations of investing in a particular stock to give huge profits, is encrypted and stored in the cloud database. The encryption is carried out to prevent the opposition from knowing the hidden stock quotes thereby ensuring security. The work is developed using 100000 tweets with the time stamps captured to encrypt and decrypt the tweets obtained after pre processing through web crawler mechanism. The overall intent of the work developed is to enable people looking at tweets for stock market quotations about investment options to proceed in a right direction, rather than ending up incurring losses due to improper assessment of tweets.