Domain-Specific Use Cases for Knowledge-Enabled Social Media Analysis

Document Type

Book Chapter

Publication Date

9-2018

Abstract

Social media provides a virtual platform for users to share and discuss their daily life, activities, opinions, health, feelings, etc. Such personal accounts readily generate Big Data marked by velocity, volume, value, variety, and veracity challenges. This type of Big Data analytics already supports useful investigations ranging from research into data mining and developing public policy to actions targeting an individual in a variety of domains such as branding and marketing, crime and law enforcement, crisis monitoring and management, as well as public and personalized health management. However, using social media to solve domain-specific problem is challenging due to complexity of the domain, lack of context, colloquial nature of language, and changing topic relevance in temporally dynamic domain. In this article, we discuss the need to go beyond data-driven machine learning and natural language processing, and incorporate deep domain knowledge as well as knowledge of how experts and decision makers explore and perform contextual interpretation. Four use cases are used to demonstrate the role of domain knowledge in addressing each challenge.

DOI

10.1007/978-3-319-94105-9_9

Link to Full Text


Share

COinS