Crowd-Labeling Fashion Reviews with Quality Control

We present a new methodology for high-quality labeling in the fashion domain with crowd workers instead of experts. We focus on the Aspect-Based Sentiment Analysis task. Our methods filter out inaccurate input from crowd workers but we preserve different worker labeling to capture the inherent high...

Full description

Saved in:
Bibliographic Details
Published in:arXiv.org 2018-04
Main Authors: Chernushenko, Iurii, Gers, Felix A, Löser, Alexander, Checco, Alessandro
Format: Article
Language:English
Subjects:
Data mining
Labeling
Quality control
Sentiment analysis
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:We present a new methodology for high-quality labeling in the fashion domain with crowd workers instead of experts. We focus on the Aspect-Based Sentiment Analysis task. Our methods filter out inaccurate input from crowd workers but we preserve different worker labeling to capture the inherent high variability of the opinions. We demonstrate the quality of labeled data based on Facebook's FastText framework as a baseline.
ISSN:2331-8422