Evaluating chatbot assistance in historical document analysis

David Hazemali (Author), Janez Osojnik (Author), Tomaž Onič (Author), Tadej Todorović (Author), Mladen Borovič (Author)

Abstract

The article explores the potential of PDFGear Copilot, a chatbot-based PDF editing tool, in assisting with the analysis of historical documents. We evaluated the chatbot's performance on a document relating to the Slovenian War of Independence. We included 25 factual and 5 interpretative questions to address its formal characteristics and content details, assess its capacity for in-depth interpretation and contextualized critical analysis, and evaluate the chatbot’s language use and robustness. The chatbot exhibited some ability to answer factual questions, even though its performance varied. It demonstrated proficiency in navigating document structure, named entity recognition, and extracting basic document information. However, performance declined significantly in tasks such as document type identification, content details, and tasks requiring deeper text analysis. For interpretative questions, the chatbot's performance was notably inadequate, failing to link cause-and-effect relationships and provide the depth and nuance required for historical inquiries.

Keywords

chatbots;historical document analyses;Yugoslavia;wars;generative;artificial intelligence;large language models;

Data

Language:	English
Year of publishing:	2024
Typology:	1.01 - Original Scientific Article
Organization:	UM FERI - Faculty of Electrical Engineering and Computer Science
UDC:	004.8:930.25
COBISS:	219271683
ISSN:	2591-0884
Views:	0
Downloads:	2
Average score:	0 (0 votes)
Metadata:

Other data

Secondary language:	Slovenian
Secondary title:	Vrednotenje pogovornega sistema kot pomočnika pri analizi zgodovinskih dokumentov
Secondary abstract:	Članek preučuje potencial pogovornega sistema PDFGear Copilot kot pomočnika pri analizi zgodovinskih dokumentov. Učinkovitost pogovornega sistema smo preizkusili na dokumentu, povezanim s slovensko osamosvojitveno vojno. Zmožnost poglobljene interpretacije, kritične analize ter jezikovno robustnost pogovornega sistema smo presojali na podlagi njegovih odgovorov na 25 faktografskih in 5 interpretativnih vprašanj, povezanih z lastnostmi in vsebino dokumenta. Pri odgovarjanju na faktografska vprašanja je bil pogovorni sistem delno uspešen. Ni pa imel težav s prepoznavanjem strukture dokumenta, razpoznavanjem imenskih entitet in pridobivanjem osnovnih informacij o dokumentu. Slabše se je odrezal pri opredelitvi vrste dokumenta, navajanju vsebinskih podrobnosti in pri nalogah, ki zahtevajo poglobljeno analizo besedila. Pri interpretativnih vprašanjih je njegova uspešnost opazno upadla, saj ni bil kos vzročno-posledičnim odnosom ter ni izkazal potrebne poglobljenosti in natančnosti, potrebni za zgodovinopisne raziskave.
Secondary keywords:	pogovorni sistemi;analiza zgodovinskih dokumentov;Jugoslavija;vojne;generativna umetna inteligenca;veliki jezikovni modeli;
Type (COBISS):	Scientific work
Pages:	str. 53-83
Volume:	ǂLetn. ǂ7
Issue:	ǂšt. ǂ2
Chronology:	2024
DOI:	10.54356/MA/2024/BIUB3010
ID:	26835036