Using machine learning for stance detection

Date

2019-01-31

Authors

Jiang, Yan, M.S. in Information Studies

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

In this report, we focus on the task of stance detection with the power of machine learning techniques to determine the relative stance between a headline and an associated article. We explore the performance of various models on the fake news challenge dataset, including both traditional methods and neural networks. We find that the Enhanced Sequential Inference Model proposed for Natural Language Inference achieves great performance on the task of stance detection. To address the imbalanced class distribution problem, we use Focal Loss to train our neural networks.

Department

Description

LCSH Subject Headings

Citation