University of Twente Student Theses

Login

Exploring business opportunities of sentiment analysis : a guideline for businesses, from implementation to business value generation

Haarhuis, L.M. (2022) Exploring business opportunities of sentiment analysis : a guideline for businesses, from implementation to business value generation.

[img] PDF
3MB
Abstract:Machine-learning conquers the world as this emerging new technology offers great opportunities not only for science but also for businesses. One interesting machine-learning application is sentiment analysis. Sentiment analysis is an interesting and upcoming machine-learning application which measures sentiment. This machine-learning method can obtain sentiment from stakeholders like customers and employees. Research papers describe that sentiment analysis is used to predict the sentiment of Twitter users (Pozzi et al., 2016). Literature also discusses several sentiment analysis applications for businesses but the full potential of sentiment analysis for business has thus far not been examined extensively. Therefore, this research provides a comprehensive guideline for businesses on if and how they should implement and utilize sentiment analysis. The first steps (step 1 to 4) within the guideline explain how businesses can implement sentiment by utilizing pre-trained sentiment analysis models. Therefore, this research first examines which input is suitable for sentiment analysis and how this data should be cleaned in order to obtain accurate sentiment analysis scores. Next, an overview of pre-trained sentiment analysis models for the Dutch language will be presented. The research examines considerations in order to provide a guideline for businesses to select an appropriate sentiment analysis model. This research utilizes interviews to help answer step 1 to 4. The up-following steps (step 5 to 8) describe how businesses can utilize a systematic approach to explore new value-generating sentiment analysis applications for their business. The systematic approach explores new sentiment analysis applications in a business setting. This method examines the goals of several departments. It develops business metrics which can be derived from sentiment analysis. After this, applications can be developed by analyzing which metrics help each division to reach their goals. Lastly, this research ranks the applications based on business value. This helps companies to determine if they should or should not invest in sentiment analysis. In step 5 to 8 the research utilizes a case study at company X to determine how well the systematic approach of exploring new sentiment analysis applications works. Company X is a big insurance company in the Netherlands and would like to utilize the full potential of sentiment analysis for their organization. The case study entails interviews and a survey to examine sentiment analysis inputs, models, considerations, goals, metrics and applications. It compares literature to the results of the interviews and surveys and determines how companies can best implement and utilize sentiment analysis. The sentiment analysis applications are categorized in the following business divisions: board of directors, business intelligence, customer relationship management, customer support, external Affairs, human resource management, marketing, purchase department and sales department. The research shows that the case study provides valuable information for companies on how to best implement and utilize sentiment analysis. New methods are explained related to sentiment analysis input, metrics and applications. The results show the importance of defining the purpose of the sentiment analysis as the input needs to fit the purpose. Next to this, the research presents newer and broader sentiment analysis metrics like shareholder sentiment KPIs and supplier sentiment KPIs. KPIs can also be more focused like measuring the sentiment within contact moments to analyze why sentiment is going up or down during contact moments. The results also present new applications. The most valuable application presented in this research is the prioritization tool. The prioritization tool creates business value by identifying which business processes create lots of negative customer sentiment. Companies should also take into consideration that there are also limitations to sentiment analysis. Companies should examine data privacy and data security legislation in order to determine if the data can be used for sentiment analysis.
Item Type:Essay (Master)
Faculty:BMS: Behavioural, Management and Social Sciences
Subject:54 computer science, 85 business administration, organizational science
Programme:Business Administration MSc (60644)
Link to this item:https://purl.utwente.nl/essays/93495
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page