Hrapko, J. (2015). Promising applications of big data and sentiment analysis [Master Thesis, Technische Universität Wien; Wirtschaftsuniversität Wien]. reposiTUm. https://doi.org/10.34726/hss.2015.34364
big data; sentiment analysis; marketing; industries
en
Abstract:
This thesis search the business value of big data for marketing especially in area of sentiment analysis. Valuable information comes from the data and the more data we have the more accurate information we get. This thesis start with description of big data and why big data become interesting in current decade. Following with listing possible data sources, ways to get the data with -big data- technologies and also pointing out the limitation of data, such as in means of privacy. Sentiment analysis is currently interesting topic for companies to mine opinions from unstructured data sources. Mining for sentiment can be challenging even with intelligent systems. Human speech is still very advanced and can pose challenge to recognize irony or simple negations. I searched and evaluated the ways to get sentiment analysis with simple program examples. By using expert interviews and various sources I searched various implementations and references. I categorized them into separate industry sectors to find a business values for potential organizations of specific type. Potential is high but not for every company, whereas big data project can pose a challenge for smaller companies.