Tackling Hallucinations in Neural Chart Summarization

Result code in IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11320%2F23%3A10476026" target="_blank" >RIV/00216208:11320/23:10476026 - isvavai.cz</a>

Result on the web

—

DOI - Digital Object Identifier

—

Result language

angličtina

Original language name

Tackling Hallucinations in Neural Chart Summarization

Original language description

Hallucinations in text generation occur when the system produces text that is not grounded in the input. In this work, we tackle the problem of hallucinations in neural chart summarization. Our analysis shows that the target side of chart summarization training datasets often contains additional information, leading to hallucinations. We propose a natural language inference (NLI) based method to preprocess the training data and show through human evaluation that our method significantly reduces hallucinations. We also found that shortening long-distance dependencies in the input sequence and adding chart-related information like title and legends improves the overall performance.

Czech name

—

Czech description

—

Type

D - Article in proceedings

CEP classification

—

OECD FORD branch

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Project

—

Continuities

R - Projekt Ramcoveho programu EK

Publication year

2023

Confidentiality

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Article name in the collection

Proceedings of the 16th International Natural Language Generation Conference

ISBN

979-8-89176-001-1

ISSN

—

e-ISSN

—

Number of pages

10

Pages from-to

414-423

Publisher name

Association for Computational Linguistics

Place of publication

Stroudsburg, PA, USA

Event location

Praha, Czechia

Event date

Sep 13, 2023

Type of event by nationality

WRD - Celosvětová akce

UT code for WoS article

—