University of Twente Student Theses

Login

Praktisch bruikbare “social media reputation” index : big social data-analyse legt social media reputation- indexmethode bloot

Tax, M.M.J.W. (2014) Praktisch bruikbare “social media reputation” index : big social data-analyse legt social media reputation- indexmethode bloot.

[img] PDF
4MB
Abstract:Development: The Social Media Reputation index method was developed by using a qualitative research design. This method was based on the reputation factors as mentioned in the Reputation Quotient model (Fombrun, Gradberg & Sever, 2000). The factors social and environmental responsibility (SER), emotional appeal (EA), products and services (PS) were defined as relevant SMR indicators of the SMR index. These indicators were then used to determine the reputation within the social media domain. By labeling messages with a sentiment and SMR indicator, it became possible to calculate the individual SMR indicator scores over a certain time and these scores were then used to determine the SMR index. To provide some explanations for the possible patterns within the SMR index, knowledge from the economic sector was used. In the stock exchanges such patterns exist much longer and they can give insights in what happens with an index. The use of trend lines or the moving average graph initially display credited trends, which could also manifest in the SMR-index. In addition, four new SMR patterns were defined. These patterns are: Crossover pattern, Explosive SMR pattern, Increased or Decreased SMR pattern and SMR-Correlation pattern. The presence of these patterns were not extensively tested, but these new patterns might lead to further research. Results: Based on the results of the SMR index it was determined on what time intervals and by which contexts the reputation of the Dutch banks was affected. With this information it was possible to perform a context and media analysis. With the results of the context and media analysis it was possible to filter the cause of a certain reputation trend. One of the results showed that a peak in the number of social media messages about the financial banks was accompanied by a decrease in the SMR index. Practical usefulness and recommendations: As this is a development study, the practical usefulness of this method is also relevant. To quickly determine this index, it is important to have closer collaborations with computer scientists. This contributes to faster implementation and testing of new social media analysis. Social scientific knowledge can serve as a basis for the development and assembly of social media analysis toolings. This is relevant because the method can only be performed manually at this stage. The software that could automatically determine the contexts and sentiments, is still in its early stage. Of course, the first steps can be put in place to automate this process. Therefore, the presence of noise and inaccuracies, due to the software-based analysis of text elements, must be taken into account. The SMR-index can best be observed at fixed intervals, so that the index can be optimally used. It is possible to use the SMR index method for smaller time intervals. However, it is advisable to perform frequently and continuous measurements. Giving explanations surrounding the reputation developments, can be made easier, if the index is determined over a greater period of time.
Item Type:Essay (Master)
Faculty:BMS: Behavioural, Management and Social Sciences
Subject:05 communication studies
Programme:Communication Studies MSc (60713)
Link to this item:http://purl.utwente.nl/essays/66318
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page