Linked Data on Geo-annotated Events and Use Cases for the Resilience of Ukraine

Manar Attar, Shuai Wang, Ronald Siebes, Eirik Kultorp, Zhisheng Huang, Tianyang Lu · 2025

Read Linked Data on Geo-annotated Events and Use Cases for the Resilience of Ukraine by Manar Attar, Shuai Wang, Ronald Siebes, Eirik Kultorp, Zhisheng Huang, Tianyang Lu (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-14762.txt.

Introduction

Manar Attar and five co-authors build a linked data set of damaging events in Ukraine to support recovery and resilience. Cities under attack need joined up facts from news and social media, yet sources clash in format and terms. Linked data methods can unify, enrich, and merge such records.

The team converted two data sets on damage from Russia's invasion, covering February twenty twenty-two to April twenty twenty-three, and added place data. An algorithm then matched reports of the same event across sources. The merged set holds about ten thousand events, spanning harm to hospitals, schools, roads, and homes. Use cases show how planners can query it for resilience work.

Catalog record

CREATORManar Attar, Shuai Wang, Ronald Siebes, Eirik Kultorp, Zhisheng Huang, Tianyang Lu
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE44 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2501-14762

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Linked Data on Geo-annotated Events and Use Cases for the Resilience of Ukraine Manar Attar, Shuai Wang, Ronald Siebes, Eirik Kultorp, Zhisheng Huang, Tianyang Lu Abstract The mission of resilience of Ukrainian cities calls for international collaboration with the scientific community to increase the quality of information by identifying and integrating information from various news and social media sources. Linked Data technology can be used to unify, enrich, and integrate data from multiple sources. In our work, we focus on datasets about damaging events in Ukraine due to Russia’s invasion between February 2022 and the end of April 2023. We convert two selected datasets to Linked Data and enrich them with additional geospatial information. Following that, we present an algorithm for the detection of identical events from different datasets. Our pipeline makes it easy to convert and enrich datasets to integrated Linked Data. The resulting dataset consists of 10K reported events covering damage to hospitals, schools, roads, residential buildings, etc. Finally, we demonstrate in use cases how our dataset can be applied to different scenarios for resilience purposes. 1 1 1 The paper is an extended version of our 2023 paper titled ‘Converting and Enriching Geo-annotated Event Data: Integrating Information for Ukraine Resilience’ [ 1 ] presented at the ACM International Conference on Advances in Geographic Information Systems (SIGSPATIAL conference). It was included in the SIGSPATIAL ’23: Proceedings of the 31st ACM International Conference on Advances in Geographic Information Systems . The use cases were presented at the BNAIC-BeNeLearn Joint International Scientific Conferences on A.I. and Machine Learning. This long paper is the result of a merge of them with some minor extensions and additional explanations. Please cite our SIGSPATIAL paper instead. Its DOI is 10.1145/3589132.3625580. Related information and resources can be found on the Linked4Resilience platform ( https://www.linked4resilience.eu ). August 24, 2026 1 Introduction The Russian-Ukrainian conflict since February 2022 has damaged civilian infrastructure, facilities, and buildings, sparking a large displacement crisis in Europe. It was reported in 2022 that, in Kharkiv alone, the war has resulted in large-scale destruction of infrastructure with an estimation of more than 1,000 buildings destroyed, among which 700 are multi-storey apartment buildings and are no longer habitable [ 2 ] . Rebuilding destroyed residential buildings, public facilities, and social infrastructure is critical for the resilience of Ukraine. It is no easy task and will require international cooperation and coordination for reconstruction, including integration and management of datasets and resources of various kinds. There are many open datasets reporting the progress of the Russian-Ukrainian conflict from various perspectives. These datasets enable researchers and analysts to gain insights into the complex situation, ultimately contributing to the development of effective strategies to protect civilians, promote peace, estimate resources for projects for resilience, etc. WikiEvents [ 3 ] consists of entries automatically curated based on Wikipedia’s Current Events portal. 2 2 2 https://en.wikipedia.org/wiki/Portal:Current_events Its NLP downstream pipeline extracts 21,275 events including around a thousand events about the Russian invasion of Ukraine. ACLED is a much larger dataset 3 3 3 https://acleddata.com/2023/03/01/war-in-ukraine-one-year-on-nowhere-safe/ , visited on 13th June, 2023. with over one million events, including around 40,000 political violence events across Ukraine [ 4 ] . This dataset is mostly dedicated to military-related events about shelling, artillery, and missile strikes. The Centre for Information Resilience (CIR) launched the Eyes on Russia (EoR) 4 4 4 https://eyesonrussia.org/ project in January 2022 with the aim of gathering and verifying media related to Russia’s invasion of Ukraine. The project’s primary objective is to provide access to verified information through a database and an interactive map, benefiting journalists, non-governmental organizations (NGOs), policymakers, and the public. The interactive map displays relevant information such as the data source, a description of the event, location coordinates, and the extent of damage caused. Furthermore, it includes a variety of properties, including the country name, province, city, coordinates, date, damage level, and source of information. The Civilian Harm in Ukraine TimeMap (CH) 5 5 5 https://ukraine.bellingcat.com/ is a similar project that provides a comprehensive record of such incidents by including source links, precise location data determined by the Global Authentication Project and Bellingcat researchers, and a brief description based on visual evidence. Its structured data can be used for further analysis and research to understand better the impact of the conflict on civilians in Ukraine. EoR and CH are the two datasets selected for this study given their similar approach in generating and representing events. Both projects focus on damage reporting and serve as important resources for those seeking accurate and verified information to aid decisions for the resilience of Ukraine. Accurate and complete documentation of the damage could benefit projects for resilience in multiple ways. Linked Data is structured data that can be interlinked with other data, which enables additional functions through semantic queries. While it is not common to use Linked Data in projects for resilience, past projects have demonstrated the use of Linked Data in decision-making in government, NGOs, and societal organizations. For example, the Brazilian government used ontologies and enriched data from various governments, resulting in a DBpedia-like Government Open Linked Data - DBGOldBr [ 5 ] . Our project explores how the transformation of event data into Linked Data facilitates an integrated and more complete description of events. For example, we consider the following event 6 6 6 The event was extracted manually based on the Twitter post https://twitter.com/KyivIndependent/status/1501218105342763020 . in CH that happened on 7th March 2022. It was reported to have “Hospital destroyed by explosion”. Its location information is “Izum, Kharkiv region” and it lacks information about the postal code. While the above-mentioned datasets were initially designed for their platforms, one can take advantage of ontologies and Linked Data technologies to provide a unique representation of entities such as cities and provinces to reduce ambiguity, which could make it easier for integration and verification, and enable interoperability with datasets in other disciplines (e.g. economic and social-/historical- data). In this paper, we attempt to convert and unify structured geo-annotated datasets. More specifically, we convert two existing geo-annotated datasets dedicated to damage reporting in Ukraine to their corresponding representation as Linked Data. We propose a pipeline for the integration of datasets and demonstrate the use of the resulting dataset. Finally, we evaluate the quality of the integrated data and demonstrate its application by several use cases. Our research question is: How to unify geo-annotated events in multiple datasets about damaging events? We answer this question by studying the following sub-research questions: SRQ1: How can we provide a unified representation of information in the datasets as Linked Data? SRQ2: How can we enrich the converted Linked Data with geospatial information? SRQ3: How can we integrate datasets by identifying and merging entities that describe the same events? SRQ4: What is the quality of the resulting unified data? The research output of the paper includes the converted dataset together with related resources, an integrated dataset, a reusable pipeline with open source code that can be adapted to future datasets, the mapping between entities in the datasets, and some use cases with their SPARQL queries. This paper is organized as follows. Section 2 investigates related approaches to the representation of damaging and resilience information. Section 3 compares

Read the complete text