Pak and Patrick (2010) defined that Microblogging currently became one of the crucial kind of the communication among Internet users. People use microblogging to talk about their daily activities and to seek or share information and show how users with similar intentions connect with each other. Jansen, et al. (2009) identified microblogging as online word-of-mouth branding in their recent research. Especially, it has been easy to reach the microblogging on every platform within seconds by expanding use of the smart phones and mobile devices applicants. By using the microblogging, people can reach valuable information sources from the different people, located in different places, about various topics. People share their thoughts and current experiences simultaneously when international or national events happen. From this perspective, they sometimes behave as a news reporter to other micro blogging users. In addition, they provide related information to users much faster than news on TV or Radio. Castillo, et al. (2011) pointed out that microblogging acquires more significance as a valid news resource, particularly for emergency situations and substantial events; it becomes critical to contribute tools’ credibility of online information. Moreover, micro blogging can behave as a public opinion poll since they are predictor tools to estimate a result of political elections, social perception and economic expectation. Hence, micro blogging is very popular for internet users in order to understand the sentiments and opinion of public about any kind of events.
By increasing the use of blogs and social networks, linguistic and sentiment analysing became a major field for many researches and businesses. Particularly, microblogging messages are notably used for marketing researches and social studies in order to understand public opinion about a specific topic as well. Yu and Kak (2012) emphasized that surveying the qualities and contents of social media provides us an opportunity to identify social structure characteristics and it also sometimes the ability to predict future human related topics. Huberman, et al. (2008) indicated that researchers, publicists and politician sight huge social network users and they used this representation of social interactions to study propagation of ideas, social bond dynamics and viral marketing.
One of the well-known micro blogging utilizes is related to stock messages. It is commonly known that price movements in the stock exchanges are formed by people’s opinions and decisions. In consideration of making a decision, people use online forums to exchange their ideas and other stock-related data. Nofsinger (2005) remarked that the communal measurement of positive and negative feelings and opinions in public arena is demonstrated by the sentiments of financial administrator because these feelings and opinions are correlated with economic participators. Moreover, the author expressed that businesses, financial professionals and consumers are relatively affected from the social mood when they identify sentiment based investment opportunities.
Index of Consumer Sentiment (O’Connor et al. 2010) BAHSET
Millions of messages that are related to the discussion of commercial trading ideas are sent to popular microblogging web-sites like Twitter, Facebook, and Tumblr per a second. Especially, Twitter, is commonly used worldwide actively, and ranks as the most popular social networks at the moment in accordance to current social media industry characters.
Twitter, the most popular microblogging service. Twitter Istatisk ver. Konum bazl? servisler
TW?TTER ??RKET B?LG?LER?N? DÖ?E, piyasa DEGER?N? GRAF?KLE VER
TW?TTER VOLUME ?LE REKLAM GEL?RLER? ARASINDAK? ?L???K?YE BAK
TW?TTER ?STAT?KLER?N? DÖ?E As of the fourth quarter of 2015, Twitter had reached 305 million monthly active users.
tweets related to Apple Inc.). TWEETERE ÖZELL?KLERDEN BAHSET
Twitter defined as an online social networking service which provides an opportunity to users print short messages, up to 140 characters, called “tweets”. Also, Twitter API1 allows users to collect message texts for scientific and statistical researches. Twitter API is easy way to collect millions of tweets.
Jansen, et al. (2009) inserted that it is apparent that micro-blogging services such as Twitter could become key applications in the attention economy.
After Bollen et al. published the article in 2010, media showed great interest in the pater and they reported it widely. Among these media, USA Today, Wired, MIT Technology Review, The Atlantic, Time, CNN, The Telegraph, CNBC etc. can be displayed as an example. In one of these media broadcast, Bollen talked about he had a patented process based on Twitter for stock forecast on Fox Business Channel in 2013.
FAKAT BUNA ?T?RAZLAR GELD? http://sellthenews.tumblr.com/post/21067996377/noitdoesnot
Jordan (2010) wrote an article on Bloomberg News Website. He reported the article by heading that “Hedge Fund Will Track Twitter to Predict Stock Moves” (Jordan, 2010). The journalist interviewed with Co-owner Paul Hawtin of The Derwent Absolute Return Fund Ltd. (Jordan, 2010). Hawtin said that they have a trading activity which is based on Twitter sentiment analyses. He talked in detail about they have a contract with some university scholars in order to write a paper about how to forecast DJIA index by using twitter sentiments.
Mackintosh interviewed with Paul Hawtin, chief executive and founder of Derwent Capital in 2012. Derwent said that they analyse the tweets and supply their special customers a system. The system produce sentiment signal in order to alert when emotionality against their securities acts precipitously (Mackintosh, 2012)
Substantially, many stock micro blogs are derived from Twitter post messages. For instance, Twitter related third-party applications such as TweetTrader.net2 and Stocktwits.com3 filter stock related Twitter posts and they serve their clients opportunity to compare stock prices, analyse message figures, sentiment evaluation, and anticipate fluctuations in securities exchange. These blogs derived from Twitter are also based on trading systems which have been developed by financial professionals because the purpose is to alert users of sentiment-based investment opportunities.
Moreover, Sprenger (2010) investigated two stock microblogging forums to be used for gathering the information including stock-related tweets. One of them is TweetTrader.net support to the erudition of crowds based on insights from scholastic research on stock microblogs; the application integrates inputs from text classification, user voting and a proprietary Stock Game in order to obtain the sentiment of online investors with respect to all publicly traded companies of the S 500 (Sprenger, 2010).
In this context, linguistic analyse and opinion extraction is going to be studied on stock related Twitter messages. Dataset collected messages from Twitter by an API.
Firstly, the relationship was investigated between Stock related Twitter messages and Dow Jones Index, then offer an explanation of the relation between DJI’s abnormal stock return, volume, volatility and micro blogging’s sentiment analyse and frequencies.
Secondly, the relationship was investigated between Stock related Twitter posts and BIST 30 Index and BIST 100 Index. Then, an explanation of the relation was offered between BIST 30’s abnormal stock return, volume, volatility and micro blogging’s sentiment analyse and frequencies.
Finally, I will try to perform whether the stock market really is predicted by Twitter subjectivity analysis.
1 Application Program Interface