PL | EN

AI.IHPAN.EDU.PL

Server with artificial intelligence tools (LLM/ML/NLP) dedicated to historical research

Logo IH PAN

Applications

⚠️ Some applications are available only for internal use by staff and collaborators of the Institute of History of the Polish Academy of Sciences.

WikiHumSearch →

Semantic (vector) data search in the WikiHum knowledge base.

SGKP-Search →

Hybrid search tool for the Geographical Dictionary of the Kingdom of Poland.

ScansAndTranscriptions →

A simple desktop scan and transcript viewer with the option to read scans using the Gemini Pro 3 model and the ability to manually correct transcriptions.

Text2NER →

Text2NER - a tool for processing text files into TEI-XML format with automatic tagging of people and locations.

MAPE Search Engine →

MAPE Search Engine - Explore the Letters That Built the Colonial Portuguese Empire. Step into history with AI search engine and chatbot

OCR Latin Check →

OCRLatinCheck - a tool for finding potential errors after OCR/HTR in Latin texts

Manuscripts Lab →

Manuscripts Lab - an application for analyzing Latin documents: quality analysis of HTR and machine translations

LegationesAdVaticanum →

LegationesAdVaticanum - Database of Polish missions to the Holy See in the 15th century

Regulations and Orders Assistant →

AI Assistant for regulations and orders of the Institute of History of the Polish Academy of Sciences

Tutorials and documentation

Introduction to eScriptorium and Kraken (HTR) - in Polish

eScriptorium is an Open Source web application designed for working with historical manuscripts and prints - preparing manual and automatic transcriptions. The application is integrated with Kraken, a tool using deep learning algorithms for text recognition (OCR and HTR). The project is developed by a team from École Pratique des Hautes Études - Université PSL.

Artificial intelligence in historical research - presentation (in Polish)

Presentation on the possibilities of using Artificial Intelligence in historical research - prepared in October 2024.

Preparing digital editions supported by natural language processing tools - presentation (in Polish)

Presentation delivered in November 2024 at the conference Data and Language: Research Perspectives in Digital Editions of Historical Sources

Building a Knowledge Graph for historical biographies using LLM

Presentation delivered at the European Social Science History Conference (ESSHC) 2025

Publications

2025

Guzik-Jureczka, Maria, Piotr Jaskulski, and Adam Zapała. "Rola Narzędzi Cyfrowych w Przetwarzaniu i Udostępnianiu Biogramów Postaci Historycznych.", Kwartalnik Historii Nauki i Techniki 70, no. 1 (2025): 39.

https://doi.org/10.4467/0023589XKHNT.25.001.21319 →

2025

Jaskulski, Piotr, Tomasz Latos, Mariusz Ryńca, and Adam Zapała. "Reliability of Large Language Models as a Tool for Knowledge Extraction from Biographical Dictionaries: The Case of the Polish Biographical Dictionary.", Digital Scholarship in the Humanities 40, no. 2 (2025): 538–48.

https://doi.org/10.1093/llc/fqaf014. →

Blog

Repositories, data and models