Event Extraction and Representation: A Case Study for the Portuguese Language

Paulo Quaresma,Kashyap Raiyani,Roy Bayot,Vítor Beires Nogueira

doi:10.3390/info10060205

Abstract

Text information extraction is an important natural language processing (NLP) task, which aims to automatically identify, extract, and represent information from text. In this context, event extraction plays a relevant role, allowing actions, agents, objects, places, and time periods to be identified and represented. The extracted information can be represented by specialized ontologies, supporting knowledge-based reasoning and inference processes. In this work, we will describe, in detail, our proposal for event extraction from Portuguese documents. The proposed approach is based on a pipeline of specialized natural language processing tools; namely, a part-of-speech tagger, a named entities recognizer, a dependency parser, semantic role labeling, and a knowledge extraction module. The architecture is language-independent, but its modules are language-dependent and can be built using adequate AI (i.e., rule-based or machine learning) methodologies. The developed system was evaluated with a corpus of Portuguese texts and the obtained results are presented and analysed. The current limitations and future work are discussed in detail.

Highlights

Text information extraction is an important natural language processing (NLP) task, aimed at automatically identifying, extracting, and representing information from text
We will focus on the research about events, focusing on two questions: What are the primitive elements of events and how can they be automatically extracted?
Regarding the evaluation of our system, we will only present the results obtained with the new modules and with the overall performance of the system

Summary

Introduction

Text information extraction is an important natural language processing (NLP) task, aimed at automatically identifying, extracting, and representing information from text. Event extraction is an important and relevant sub-task in the NLP domain [1]. The conventional view of events is that, given a sentence, events denote an activity or a state of action. The extracted information can be represented by specialized ontologies [2], supporting knowledge-based reasoning and inference processes. This topic has gained relevance with the exponential growth of social networks and the need to automatically identify and extract referred events [3,4]. We will focus on the research about events, focusing on two questions: What are the primitive elements of events and how can they be automatically extracted?

Objectives

Results

Discussion

Conclusion