The stock market is fluctuating constantly. The
rise and fall in stock prices are seemingly random. However,
this is not so. Even a minute happening in the company can
have a huge effect on the stock price. As each investor buys
and sells the stock, the price rises and falls depending on the
sale and purchase, the demand and supply. Whether or not an
investor buys a particular company's stock is based on his
knowledge and impression of the company. The latter is what
we will employ to decide whether or not to buy a certain
company's stock at the current price. There are 6 accepted
discrete moods. Millions of people tweet every second. A
fairly accurate prediction and analysis of the tweet's underlying
mood can be made using sentiment analysis. Each word has a
certain grammatical signature that tells us which mood it
belongs to. Depending on what the users are feeling about a
company as they tweet about it this engine will decide whether
or not one should buy stocks of that company.
This paper describes how to map this mood with market
sentiment and in turn with prediction of rise/fall of stock
prices.
Tejas Sathe : Dept. of Computer Engineering
MIT College of Engineering
Paud Road, Pune.
Siddhartha Gupta : Dept. of Computer Engineering
MIT College of Engineering
Paud Road, Pune.
Shreya Nair : Dept. of Computer Engineering
MIT College of Engineering
Paud Road, Pune.
Sukhada Bhingarkar : Dept. of Computer Engineering
MIT College of Engineering
Paud Road, Pune.
Twitter
Sentiment Analysis
Successive
Deviations
Stock Market
Mood
We successfully extracted tweets and performed
sentiment analysis on them. They give a fairly accurate
guess of the direction of stock market shift. The
recommendations generated based on clustering and
ranking are quite sound based on our manual monitoring
The entire processing takes a lot of time. So, for
demonstration purposes we extract only 100 of the most
recent tweets. But for increasing accuracy we need much
more than that. So what could be done is distribute the
set of tweets over parallel processors and combine the
results at the end. Market sentiment is only one part of
what decides rise and fall of stock prices. So in some
cases our predictions can go wildly wrong. To improve
upon this some Business Analytics have to be added, so
that in combination with market sentiment can help the
user make a more informed decision.
[1] Sentiment Analysis:
http://en.wikipedia.org/wiki/Sentiment_analysis
[2] What is a Bull and a Bear
market?http://content.moneyinstructor.com/693/whatbull-
bear-market.html
[3] The Role of the Stock Market:
http://www.updown.com/education/article/The-Roleof-
the-Stock-Market
[4] Daily Market Summary:
http://www.nasdaqtrader.com/Trader.aspxid=DailyMar
ketSummary
[5] Ray Chen, Marius Lazer, “Sentiment Analysis of
Twitter Feeds for the Prediction of Stock Market
Movement”
[6] SentiWordNet. An Enhanced Lexical Resource for
Sentiment Analysis and Opinion Mining.
http://sentiwordnet.isti.cnr.it/
[7] Alexander Pak, Patrick Paroubek, “Twitter as a Corpus
for Sentiment Analysis and Opinion Mining”,
Proceedings of the Seventh conference on International
Language Resources and Evaluation LREC'10,
Valletta, Malta, European Language Resources
Association ELRA,(May 2010)
[8] Namrata Godbole, Manjunath Srinivasaiah, Steven
Skiena, “Large-Scale Sentiment Analysis for News and
Blogs”, Proceedings of the International Conference
on Weblogs and Social Media ICWSM, (2007)
[9] Efthymios Kouloumpis, Theresa Wilson, Johanna
Moore, “Twitter Sentiment Analysis: The Good the
Bad and the OMG!”, Proceedings of the International
Conference on Weblogs and Social Media ICWSM
(2011).
