Citation
Gascó Sánchez, Luis and Asensio Rivera, César and Arcas Castro, Guillermo de and Clavel, Chloé
(2019).
Noytext: A Web platform to annotate social media documents on noise perception for their use in opinion mining research.
In: "Inter Noise 2019: Noise control for a better environment", 16 al 19 de junio de 2019, Madrid (España). pp. 1-8.
Abstract
Boost of online social networks has demonstrated that some people are willingto share their views about everyday problems, including noise. With the advent ofNatural Language Processing and Machine Learning technologies to the majority ofthe scientific fields, we have begun to analyze the textual content of social media, andmore specifically online social networks, to extract insights about the noise attitudeof the population that uses this channel to express their opinion in this matter.Some of the state-of-the-art algorithms, such as deep neural networks, aresupervised machine learning algorithms.This means that researchers have toprovide a set of labelled training data to build new models. The annotation processis known as one of the most time-costly tasks in a data science pipeline, sinceresearchers among other thigs have to test the agreement between annotators and tomeasure the quality of the categories they had previously defined. For that reasonin this paper, we introduce Noytext which is a customizable web tool to annotatetexts from your database, that can be deployed in your own webserver and you canuse to request help from colleagues and collaborators in the annotation process in afriendly way.