Vinsamlegast notið þetta auðkenni þegar þið vitnið til verksins eða tengið í það: http://hdl.handle.net/1946/22592
With the increased use of social media platforms, being present on these platforms is not sufficient for companies to succeed and be ahead of their competitors. Nowadays, a large portion of consumers demand effective information flow from companies on social media and expect to receive sufficient amount of appropriate information by exchanging a very limited amount of effort. However, companies need to make sure that their social media messages are visible to as many people as possible, and also that the information the messages contain is of value to those specific customers. That being said, it is important to identify what kind of messages attract the most attention and, especially when it comes to Twitter, influential users to engage in the discussion.
In this paper, it is researched whether Tibco Spotfire can be effectively utilized in order to identify these influential Twitter users and other elements singled out by the theories under discussion as well as an additional measure is introduced, which can be implemented to increase the validity of the results. A dataset provided by Icelandair that includes every single tweet posted concerning the company between September 2013 and March 2015 is used to test the software and realize its capabilities. However, the dataset has its flaws due to the method used to extract it and therefore limits the research as it fails to provide accurate and reliable data for two of the nine variables of interest. Nonetheless, the dataset is able to assist in realizing the usefulness and potential of the software.
The results show that Tibco Spotfire can be used effectively to enhance social media performance as it manages to identify the elements put forth by the two social media theories under discussion. The software manages to identify the most influential users, both those with the largest follower base and the ones that have been the most active ones as well as it manages to identify several other variables put forth by the theories. However, there are several limitations to the research such as the inaccurate database, limited experience and knowledge with the software and limited amount of theories taken into consideration.