Knowledge Graphs and Natural-Language Processing

TL;DR

This paper discusses how knowledge graphs and semantic technologies can enhance emergency data analysis, especially from natural language sources like social media, by providing efficient, flexible, and standardized data management tools.

Contribution

It provides an overview of semantic technologies supporting knowledge graphs and discusses their application to natural language data in emergency management.

Findings

Knowledge graphs offer a flexible data representation for emergencies.

Semantic technologies facilitate integration of diverse data sources.

Natural language processing techniques are crucial for social media data analysis.

Abstract

Emergency-relevant data comes in many varieties. It can be high volume and high velocity, and reaction times are critical, calling for efficient and powerful techniques for data analysis and management. Knowledge graphs represent data in a rich, flexible, and uniform way that is well matched with the needs of emergency management. They build on existing standards, resources, techniques, and tools for semantic data and computing. This chapter explains the most important semantic technologies and how they support knowledge graphs. We proceed to discuss their benefits and challenges and give examples of relevant semantic data sources and vocabularies. Natural-language texts -- in particular those collected from social media such as Twitter -- is a type of data source that poses particular analysis challenges. We therefore include an overview of techniques for processing natural-language texts.