Paper Type
Research-in-Progress Paper
Description
Social Media have emerged as an additional interaction channel for companies with their partners, employees and customers. However, the pace of interaction in this channel is as high as the data volume that may be relevant for businesses. Therefore, mechanisms other than manually monitoring and analyzing this data are needed. Basic approaches have emerged in the areas of Social Search and Social Media Monitoring, but searches based on keywords and simple grammar are limited in their results and deliver only first indications. The main challenge with more sophisticated approaches in the domain of ontology engineering is that they require considerable investments to establish domain-specific ontologies which has often prevented their use in many cases for Social Media Analysis. This research presents work in progress and suggests an approach which aims at increasing the efficiency of defining ontologies by automatically extracting knowledge from existing enterprise application systems. For this purpose, established ontology engineering approaches are evaluated and combined with a text mining tool. The ontology serves as a "dictionary" for the analysis of unstructured Social Media content and contributes to an efficient and adaptive monitoring and analysis of unstructured content.
TOWARDS AN ONTOLOGY-BASED APPROACH FOR SOCIAL MEDIA ANALYSIS
Social Media have emerged as an additional interaction channel for companies with their partners, employees and customers. However, the pace of interaction in this channel is as high as the data volume that may be relevant for businesses. Therefore, mechanisms other than manually monitoring and analyzing this data are needed. Basic approaches have emerged in the areas of Social Search and Social Media Monitoring, but searches based on keywords and simple grammar are limited in their results and deliver only first indications. The main challenge with more sophisticated approaches in the domain of ontology engineering is that they require considerable investments to establish domain-specific ontologies which has often prevented their use in many cases for Social Media Analysis. This research presents work in progress and suggests an approach which aims at increasing the efficiency of defining ontologies by automatically extracting knowledge from existing enterprise application systems. For this purpose, established ontology engineering approaches are evaluated and combined with a text mining tool. The ontology serves as a "dictionary" for the analysis of unstructured Social Media content and contributes to an efficient and adaptive monitoring and analysis of unstructured content.