Abstract

Social media posts created by customers capture a lot of business relevant information for decision-makers, e.g., current consumer expectations on products and services. For that purpose, the social media posts need to be analyzed thoroughly. In this respect, a topic-related classification facilitates managerial decision-making because business relevant topics, social media users discuss about, immediately become obvious and the need for action can be derived. For instance, it may get obvious that the majority of a company’s negative customer posts refers to a particular product or a specific campaign. However, such a classification of social media posts is particularly challenging for small and medium-sized enterprises (SMEs). This is because human resources for a manual examination of posts are missing and an automatic analysis is error-prone due to particularities of customer posts such as the occurrence of regional dialect or branch-specific expressions. We thus develop a tool, which enables the automatized topic-related classification of social media posts and matches the particular requirements of SMEs in southern Germany. Our solution is evaluated by using a data set stemming from three collaborating companies.

First Page

2034

Last Page

2050

Share

COinS