Persuasion techniques are powerful rhetorical devices used to sway public opinion in a wide range of media. We present a new corpus of persuasion techniques, focusing on Slavic languages. The corpus contains documents in Bulgarian, Polish, and Russian, annotated with persuasion techniques at the coarse-grained text-span level and fine-grained sentence level. The techniques are drawn from a taxonomy of 25 fine-grained persuasion techniques, grouped under six broad categories of rhetorical persuasion strategies. The corpus contains approximately 7500 text spans from 222 documents that cover topics hotly debated at the national and international levels. We describe the corpus creation process, provide detailed statistics, and examine correlations between topics and persuasion techniques. We use classic ML-based and generative AI-based models to provide baselines and benchmark results for the detection and classification of persuasion techniques at the text-span level and sentence level.

Nikolaos Nikolaidis, Nicolas Stefanovitch, Purificação Silvano, Dimitar Iliyanov Dimitrov, Roman Yangarber, Nuno Guimarães, Elisa Sartori, Ion Androutsopoulos, Preslav Nakov, Giovanni Da San Martino, Jakub Piskorski. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2025.

We systematically explore the predictive power of features derived from Persuasion Techniques detected in texts, for solving different tasks of interest for media analysis; notably: detecting mis/disinformation, fake news, propaganda, partisan news and conspiracy theories. Firstly, we propose a set of meaningful features, aiming to capture the persuasiveness of a text. Secondly, we assess the discriminatory power of these features in different text classification tasks on 8 selected datasets from the literature using two metrics. We also evaluate the per-task discriminatory power of each Persuasion Technique and report on different insights. We find out that most of these features have a noticeable potential to distinguish conspiracy theories, hyperpartisan news and propaganda, while we observed mixed results in the context of fake news detection.

This paper presents a corpus manually annotated with named entities for six Slavic languages — Bulgarian, Czech, Polish, Slovenian, Russian, and Ukrainian. This work is the result of a series of shared tasks, conducted in 2017–2023 as a part of the Workshops on Slavic Natural Language Processing. The corpus consists of 5,017 documents on seven topics. The documents are annotated with five classes of named entities. Each entity is described by a category, a lemma, and a unique cross-lingual identifier. We provide two train-tune dataset splits — single topic out and cross topics. For each split, we set benchmarks using a transformer-based neural network architecture with the pre-trained multilingual models — XLM-RoBERTa-large for named entity mention recognition and categorization, and mT5-large for named entity lemmatization and linking.

Jakub Piskorski, Nicolas Stefanovitch, Nikolaos Nikolaidis, Giovanni Da San Martino, Preslav Nakov. Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2023.

In this paper we report on the complexity of persuasion technique annotation in the context of a large multilingual annotation campaign involving 6 languages and approximately 40 annotators.We highlight the techniques that appear to be difficult for humans to annotate and elaborate on our findings on the causes of this phenomenon.We introduce Holistic IAA, a new word embedding-based annotator agreement metric and we report on various experiments using this metric and its correlation with the traditional Inter Annotator Agreement (IAA) metrics.However, given somewhat limited and loose interaction between annotators, i.e., only a few annotators annotate the same document subsets, we try to devise a way to assess the coherence of the entire dataset and strive to find a good proxy for IAA between annotators tasked to annotate different documents and in different languages, for which classical IAA metrics can not be applied.

This paper presents, to the best of our knowledge, the first ever publicly available annotated dataset for sentiment classification and semantic polarity dictionary for Georgian. The characteristics of these resources and the process of their creation are described in detail. The results of various experiments on the performance of both lexicon- and machine learning-based models for Georgian sentiment classification are also reported. Both 3-label (positive, neutral, negative) and 4-label settings (same labels + mixed) are considered. The machine learning models explored include, i.a., logistic regression, SVMs, and transformed-based models. We also explore transfer learning- and translation-based (to a well-supported language) approaches. The obtained results for Georgian are on par with the state-of-the-art results in sentiment classification for well studied languages when using training data of comparable size.

We introduce a new set of benchmark datasets derived from ACLED data for fine-grained event classification and compare the performance of various state-of-the-art machine learning models on these datasets, including SVM based on TF-IDF character n-grams and neural context-free embeddings (GLOVE and FASTTEXT) as well as deep learning-based BERT with its contextual embeddings.The best results in terms of micro (94.3-94.9%)and macro F 1 (86.0-88.9%)were obtained using BERT transformer, with simpler TF-IDF character n-gram based SVM being an interesting alternative.Further, we discuss the pros and cons of the considered benchmark models in terms of their robustness and the dependence of the classification performance on the size of training data.

In the two recent decades various security authorities around the world acknowledged the importance of exploiting the ever-growing amount of information published on the web on various types of events for early detection of certain threats, situation monitoring and risk analysis. Since the information related to a particular real-world event might be scattered across various sources and mentioned on different dates, an important task is to link together all event mentions that are interrelated. This article studies the application of various statistical and machine learning techniques to solve a new application-oriented variation of the task of event pair relatedness classification, which merges different fine-grained event relation types reported elsewhere into one concept. The task focuses on linking event templates automatically extracted from online news by an existing event extraction system, which contain only short text snippets, and potentially erroneous and incomplete information. Results of exploring the performance of shallow learning methods such as decision tree-based random forest and gradient boosted tree ensembles (XGBoost) along with kernel-based support vector machines (SVM) are presented in comparison to both simpler shallow learners as well as a deep learning approach based on long short-term memory (LSTM) recurrent neural network. Our experiments focus on using linguistically lightweight features (some of which not reported elsewhere) which are easily portable across languages. We obtained F1 scores ranging from 92% (simplest shallow learner) to 96.4% (LSTM-based recurrent neural network) evaluated on a newly created event linking corpus.

This paper reports on an ongoing development of a tool for extracting structured information on events a given target entity participated in from massive collections of textual documents and anchoring these events on a timescale. An overview of the current version of the tool and the underlying timeline extraction process is given. Some evaluation figures that reflect system output quality are provided too. The paper will be accompanied by a live demo of the timeline extraction tool.

An ever-growing amount of information relevant for early detection of certain threats can be extracted from on-line news. This led to an emergence of news mining tools to help analysts to digest the overflow of information and to extract valuable knowledge from on line news sources. This paper gives an overview of the fully operational Real-time News Event Extraction Framework developed for Frontex, the EU Border Agency, to facilitate the process of extracting structured information on border security-related events from on-line news. In particular, a hybrid event extraction system has been constructed, which is applied to the stream of news articles continuously gathered and pre-processed by the Europe Media Monitor - a large-scale multilingual news aggregation engine. The framework consists also of an earth browser, in which events are visualized and an event moderation tool, which allows to access the database of automatically extracted event descriptions and to clean, validate, group, enhance and export them into other knowledge repositories.

OSINT (Open-Source Intelligence) tool is constructed for Frontex to automate the process of extracting structured knowledge from on-line news articles on border-security related event at the EU borders and in the related third countries. Incidents and developments are relevant context for illegal migration. European Commission has identified and acknowledged the need for strengthening capabilities for tracking the security situation.

This paper describes ongoing work to construct a knowledge base of politically motivated violent events (PMVE) consisting of a domain ontology and of instance data accessible via a browser and visualization interface. The instance data is semi-automatically extracted from news articles gathered by the online news monitoring system of the European Commission (EMM). There are important interrelations in seemingly unrelated violent events recorded in online news articles scattered through the Web. This information can essentially be re-discovered by gathering and mapping it to a well defined, logical model. This knowledgebase can then deploy an inference mechanism resulting in the discovery of hidden relationships and facts.

We present an architecture for the integration of shallow and deep NLP components which is aimed at flexible combination of different language technologies for a range of practical current and future applications. In particular, we describe the integration of a high-level HPSG parsing system with different high-performance shallow components, ranging from named entity recognition to chunk parsing and shallow clause recognition. The NLP components enrich a representation of natural language text with layers of new XML meta-information using a single shared data structure, called the text chart. We describe details of the integration methods, and show how information extraction and language checking applications for realworld German text benefit from a deep grammatical analysis.

In this article we present SMES–SPPC, a high–performance system for intelligent extraction of structured data from free text documents. SMES–SPPC consists of a set of domain–adaptive shallow core components that are realized by means of cascaded weighted finite–state machines and generic dynamic tries. The system has been fully implemented for German; it includes morphological and on–line compound analysis, efficient POS–filtering, high–performance named–entity recognition and chunk parsing based on a novel divide–and–conquer strategy. The whole approach proved to be very useful for processing free word order languages such as German. SMES–SPPC has a good performance (more than 6000 words per second on standard PC environments) and achieves high linguistic coverage, especially for the divide–and–conquer parsing strategy, where we obtained an f–measure of 87.14% on unseen data.