Publications
Peer-reviewed papers with code, data, and slides.
2026
-
Mykola Trokhymovych, Ricardo Baeza-Yates, Alessandro Flammini, and 2 more authorsIn Proceedings of the 35th ACM International Conference on Information and Knowledge Management (CIKM ’26), 2026The convergence of large language models and social bots allows malicious actors to manipulate the information ecosystem by generating human-like content at scale. Existing models for detecting AI-generated content often fail in the wild, primarily due to the lack of ground-truth data. We address this gap through an adversarial methodology that models the impersonation of real social media users by malicious actors. Using this methodology, we curate a multilingual, cross-platform dataset of paired human and AI-generated messages. Training on such adversarial data yields accurate detection of AI-generated text. Our approach significantly outperforms existing models for content-based bot detection in real-world, out-of-distribution data.
@inproceedings{trokhymovych2026adversarial, title = {Adversarial Creation and Detection of {AI}-Generated Social Bot Content}, author = {Trokhymovych, Mykola and Baeza-Yates, Ricardo and Flammini, Alessandro and Saez-Trumper, Diego and Menczer, Filippo}, booktitle = {Proceedings of the 35th ACM International Conference on Information and Knowledge Management (CIKM '26)}, year = {2026}, url = {https://arxiv.org/abs/2606.07219}, } -
Aitolkyn Baigutanova, Francisco Navas, Pablo Aragón, and 5 more authorsIn Proceedings of the ACM Web Conference 2026, 2026Wikipedia is a critical source of information for millions of users across the Web. It serves as a key resource for large language models, search engines, question-answering systems, and other Web-based applications. In Wikipedia, content needs to be verifiable, meaning that readers can check that claims are backed by references to reliable sources. This depends on manual verification by editors, an effective but labor-intensive process, especially given the high volume of daily edits. To address this challenge, we introduce a multilingual machine learning system to assist editors in identifying claims requiring citations. Our approach is tested in 10 language editions of Wikipedia, outperforming existing benchmarks for reference need assessment. We not only consider machine learning evaluation metrics but also system requirements, allowing us to explore the trade-offs between model accuracy and computational efficiency under real-world infrastructure constraints. We deploy our system in production and release data and code to support further research.
@inproceedings{10.1145/3774904.3792803, author = {Baigutanova, Aitolkyn and Navas, Francisco and Aragón, Pablo and Trokhymovych, Mykola and Aslam, Muniza and Chou, Ai-Jou and Redi, Miriam and Sáez Trumper, Diego}, title = {Multilingual Reference Need Assessment System for Wikipedia}, year = {2026}, isbn = {9798400723070}, publisher = {Association for Computing Machinery}, address = {New York, NY, USA}, url = {https://doi.org/10.1145/3774904.3792803}, doi = {10.1145/3774904.3792803}, booktitle = {Proceedings of the ACM Web Conference 2026}, pages = {8191--8199}, numpages = {9}, location = {United Arab Emirates}, series = {WWW '26}, } -
An End-to-End Ukrainian RAG for Local Deployment. Optimized Hybrid Search and Lightweight GenerationMykola Trokhymovych, Yana Oliinyk, and Nazarii NyzhnykIn Proceedings of the Fifth Ukrainian Natural Language Processing Conference (UNLP 2026), May 2026This paper presents a highly efficient Retrieval-Augmented Generation (RAG) system built specifically for Ukrainian document question answering, which achieved 2nd place in the UNLP 2026 Shared Task. Our solution features a custom two-stage search pipeline that retrieves relevant document pages, paired with a specialized Ukrainian language model fine-tuned on synthetic data to generate accurate, grounded answers. Finally, we compress the model for lightweight deployment. Evaluated under strict computational limits, our architecture demonstrates that high-quality, verifiable AI question answering can be achieved locally on resource-constrained hardware without sacrificing accuracy.
@inproceedings{trokhymovych-etal-2026-end, title = {An End-to-End {U}krainian {RAG} for Local Deployment. Optimized Hybrid Search and Lightweight Generation}, author = {Trokhymovych, Mykola and Oliinyk, Yana and Nyzhnyk, Nazarii}, editor = {Romanyshyn, Mariana}, booktitle = {Proceedings of the Fifth {U}krainian Natural Language Processing Conference ({UNLP} 2026)}, month = may, year = {2026}, address = {Lviv, Ukraine}, publisher = {Association for Computational Linguistics}, url = {https://aclanthology.org/2026.unlp-1.19/}, pages = {223--229}, isbn = {979-8-89176-359-3}, }
2025
-
Mykola Trokhymovych, Lydia Pintscher, Ricardo Baeza-Yates, and 1 more authorIn Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 6: Industry Track), Jul 2025We introduce a next-generation vandalism detection system for Wikidata, one of the largest open-source structured knowledge bases on the Web. Wikidata is highly complex: its items incorporate an ever-expanding universe of factual triples and multilingual texts. While edits can alter both structured and textual content, our approach converts all edits into a single space using a method we call Graph2Text. This allows for evaluating all content changes for potential vandalism using a single multilingual language model. This unified approach improves coverage and simplifies maintenance. Experiments demonstrate that our solution outperforms the current production system. Additionally, we are releasing the code under an open license along with a large dataset of various human-generated knowledge alterations, enabling further research.
@inproceedings{trokhymovych-etal-2025-graph, title = {Graph-Linguistic Fusion: Using Language Models for {W}ikidata Vandalism Detection}, author = {Trokhymovych, Mykola and Pintscher, Lydia and Baeza-Yates, Ricardo and S{\'a}ez Trumper, Diego}, editor = {Rehm, Georg and Li, Yunyao}, booktitle = {Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 6: Industry Track)}, month = jul, year = {2025}, address = {Vienna, Austria}, publisher = {Association for Computational Linguistics}, url = {https://aclanthology.org/2025.acl-industry.21/}, doi = {10.18653/v1/2025.acl-industry.21}, pages = {284--294}, isbn = {979-8-89176-288-6}, } - Kateryna Akhynko, Oleksandr Kosovan, and Mykola TrokhymovychIn Proceedings of the Fourth Ukrainian Natural Language Processing Workshop (UNLP 2025), Jul 2025
This paper presents one of the top-performing solutions to the UNLP 2025 Shared Task on Detecting Manipulation in Social Media. The task focuses on detecting and classifying rhetorical and stylistic manipulation techniques used to influence Ukrainian Telegram users. For the classification subtask, we fine-tuned the Gemma 2 language model with LoRA adapters and applied a second-level classifier leveraging meta-features and threshold optimization. For span detection, we employed an XLM-RoBERTa model trained for multi-target, including token binary classification. Our approach achieved 2nd place in classification and 3rd place in span detection.
@inproceedings{akhynko-etal-2025-hidden, title = {Hidden Persuasion: Detecting Manipulative Narratives on Social Media During the 2022 {R}ussian Invasion of {U}kraine}, author = {Akhynko, Kateryna and Kosovan, Oleksandr and Trokhymovych, Mykola}, editor = {Romanyshyn, Mariana}, booktitle = {Proceedings of the Fourth Ukrainian Natural Language Processing Workshop (UNLP 2025)}, month = jul, year = {2025}, address = {Vienna, Austria (online)}, publisher = {Association for Computational Linguistics}, url = {https://aclanthology.org/2025.unlp-1.19/}, doi = {10.18653/v1/2025.unlp-1.19}, pages = {194--202}, isbn = {979-8-89176-269-5}, } -
Mykola Trokhymovych, Oleksandr Kosovan, Nathan Forrester, and 3 more authorsProceedings of the International AAAI Conference on Web and Social Media, Jun 2025Wikipedia is powered by MediaWiki, a free and open-source software that is also the infrastructure for many other wiki-based online encyclopedias. These include the recently launched website Ruwiki, which has copied and modified the original Russian Wikipedia content to conform to Russian law. To identify practices and narratives that could be associated with different forms of knowledge manipulation, this article presents an in-depth analysis of this Russian Wikipedia fork. We propose a methodology to characterize the main changes with respect to the original version. The foundation of this study is a comprehensive comparative analysis of more than 1.9M articles from Russian Wikipedia and its fork. Using meta-information and geographical, temporal, categorical, and textual features, we explore the changes made by Ruwiki editors. Furthermore, we present a classification of the main topics of knowledge manipulation in this fork, including a numerical estimation of their scope. This research not only sheds light on significant changes within Ruwiki, but also provides a methodology that could be applied to analyze other Wikipedia forks and similar collaborative projects.
@article{trokhymovych-etal-2025-characterizing, author = {Trokhymovych, Mykola and Kosovan, Oleksandr and Forrester, Nathan and Aragón, Pablo and Saez-Trumper, Diego and Baeza-Yates, Ricardo}, title = {Characterizing Knowledge Manipulation in a Russian Wikipedia Fork}, journal = {Proceedings of the International AAAI Conference on Web and Social Media}, volume = {19}, number = {1}, year = {2025}, month = jun, pages = {1924--1936}, doi = {10.1609/icwsm.v19i1.35910}, url = {https://ojs.aaai.org/index.php/ICWSM/article/view/35910}, }
2024
-
Mykola Trokhymovych, Indira Sen, and Martin GerlachIn Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), Aug 2024With over 60M articles, Wikipedia has become the largest platform for open and freely accessible knowledge. While it has more than 15B monthly visits, its content is believed to be inaccessible to many readers due to the lack of readability of its text. However, previous investigations of the readability of Wikipedia have been restricted to English only, and there are currently no systems supporting the automatic readability assessment of the 300+ languages in Wikipedia. To bridge this gap, we develop a multilingual model to score the readability of Wikipedia articles. To train and evaluate this model, we create a novel multilingual dataset spanning 14 languages, by matching articles from Wikipedia to simplified Wikipedia and online children encyclopedias. We show that our model performs well in a zero-shot scenario, yielding a ranking accuracy of more than 80% across 14 languages and improving upon previous benchmarks. These results demonstrate the applicability of the model at scale for languages in which there is no ground-truth data available for model fine-tuning. Furthermore, we provide the first overview on the state of readability in Wikipedia beyond English.
@inproceedings{trokhymovych-etal-2024-open, title = {An Open Multilingual System for Scoring Readability of {W}ikipedia}, author = {Trokhymovych, Mykola and Sen, Indira and Gerlach, Martin}, editor = {Ku, Lun-Wei and Martins, Andre and Srikumar, Vivek}, booktitle = {Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)}, month = aug, year = {2024}, address = {Bangkok, Thailand}, publisher = {Association for Computational Linguistics}, url = {https://aclanthology.org/2024.acl-long.342}, doi = {10.18653/v1/2024.acl-long.342}, pages = {6296--6311}, }
2023
-
Mykola Trokhymovych, Muniza Aslam, Ai-Jou Chou, and 2 more authorsIn Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, Aug 2023This paper presents a novel design of the system aimed at supporting the Wikipedia community in addressing vandalism on the platform. To achieve this, we collected a massive dataset of 47 languages, and applied advanced filtering and feature engineering techniques, including multilingual masked language modeling to build the training dataset from human-generated data. The performance of the system was evaluated through comparison with the one used in production in Wikipedia, known as ORES. Our research results in a significant increase in the number of languages covered, making Wikipedia patrolling more efficient to a wider range of communities. Furthermore, our model outperforms ORES, ensuring that the results provided are not only more accurate but also less biased against certain groups of contributors.
@inproceedings{10.1145/3580305.3599823, author = {Trokhymovych, Mykola and Aslam, Muniza and Chou, Ai-Jou and Baeza-Yates, Ricardo and Saez-Trumper, Diego}, title = {Fair Multilingual Vandalism Detection System for Wikipedia}, year = {2023}, isbn = {9798400701030}, publisher = {Association for Computing Machinery}, address = {New York, NY, USA}, url = {https://doi.org/10.1145/3580305.3599823}, doi = {10.1145/3580305.3599823}, booktitle = {Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining}, pages = {4981--4990}, numpages = {10}, location = {Long Beach, CA, USA}, series = {KDD '23}, } -
Mykola Trokhymovych, and Oleksandr KosovanIn Data Science and Algorithms in Systems, Aug 2023People generate vast amounts of data that can be used for analytics, data-driven decision-making, and forecasting. However, to extract value from data, we need to apply specific methods of cleaning and prepossessing it. In this paper, we observe the problem of geospatial data de-duplication, propose and implement end-to-end solutions for social-media-based data de-duplication. We apply advanced geospatial, natural language processing, and classical machine learning methods for our solution. Our tool shows high competitiveness in observed competition and can process a vast amount of data with limited computational resources.
@inproceedings{10.1007/978-3-031-21438-7_60, author = {Trokhymovych, Mykola and Kosovan, Oleksandr}, editor = {Silhavy, Radek and Silhavy, Petr and Prokopova, Zdenka}, title = {GeoDD: End-to-End Spatial Data De-duplication System}, booktitle = {Data Science and Algorithms in Systems}, year = {2023}, publisher = {Springer International Publishing}, address = {Cham}, pages = {717--727}, isbn = {978-3-031-21438-7}, url = {https://link.springer.com/chapter/10.1007/978-3-031-21438-7_60}, doi = {10.1007/978-3-031-21438-7_60}, }
2022
-
Mykola Trokhymovych, and Diego Saez-TrumperIn Wiki Workshop, Aug 2022Fact verification has become an essential task, being used in various areas from checking auto-generated content to fighting disinformation in hybrid wars. However, even though there has been relevant advances in creating automatic fact-checking systems, nowadays manual work is crucial to deliver good quality results. The manual fact verification usually consists of information retrieval, and logical reasoning for making the final verdict. In this work we concentrate on the process of searching for the fact proofs, reveal possible problems while searching, and propose a tool that helps increase the speed of fact checking without sacrificing its quality.
@inproceedings{trokhymovych2023wikifactfind, author = {Trokhymovych, Mykola and Saez-Trumper, Diego}, title = {WikiFactFind: Semi-automated fact-checking based on Wikipedia}, booktitle = {Wiki Workshop}, year = {2022}, url = {https://wikiworkshop.org/2022/papers/WikiWorkshop2022_paper_21.pdf}, }
2021
-
Mykola Trokhymovych, and Diego Saez-TrumperIn Proceedings of the 30th ACM International Conference on Information & Knowledge Management, Aug 2021With the growth of fake news and disinformation, the NLP community has been working to assist humans in fact-checking. However, most academic research has focused on model accuracy without paying attention to resource efficiency, which is crucial in real-life scenarios. In this work, we review the State-of-the-Art datasets and solutions for Automatic Fact-checking and test their applicability in production environments. We discover overfitting issues in those models, and we propose a data filtering method that improves the model’s performance and generalization. Then, we design an unsupervised fine-tuning of the Masked Language models to improve its accuracy working with Wikipedia. We also propose a novel query enhancing method to improve evidence discovery using the Wikipedia Search API. Finally, we present a new fact-checking system, the WikiCheck API that automatically performs a facts validation process based on the Wikipedia knowledge base. It is comparable to SOTA solutions in terms of accuracy and can be used on low-memory CPU instances.
@inproceedings{10.1145/3459637.3481961, author = {Trokhymovych, Mykola and Saez-Trumper, Diego}, title = {WikiCheck: An End-to-End Open Source Automatic Fact-Checking API Based on Wikipedia}, year = {2021}, isbn = {9781450384469}, publisher = {Association for Computing Machinery}, address = {New York, NY, USA}, url = {https://doi.org/10.1145/3459637.3481961}, doi = {10.1145/3459637.3481961}, booktitle = {Proceedings of the 30th ACM International Conference on Information \& Knowledge Management}, pages = {4155--4164}, numpages = {10}, location = {Virtual Event, Queensland, Australia}, series = {CIKM '21}, }
