Artificial Intelligence (AI) is a transformative force driving innovation, yet tracking AI-related advancements remains challenging due to the rapid pace of development and unstructured data from platforms like GitHub. This paper proposes an AI-driven approach to innovation detection, leveraging GitHub as a data source to systematically identify and link AI projects to organizations. Key contributions include a domain-specific taxonomy comprising 7,490 AI topics, a modular pipeline for semantic annotation and entity linking, and a trend detection framework based on Singular Spectrum Analysis (SSA). A knowledge graph is constructed to represent relationships among AI topics, projects, and companies, thereby enabling structured innovation tracking. The approach addresses challenges such as data sparsity and noise, demonstrating strengths in semantic annotation and topic categorization. Results highlight the potential for accurately detecting AI innovations and linking them to organizational entities, offering valuable insights for researchers, companies, and policymakers. This work contributes a scalable, automated approach for AI innovation tracking, with future directions focusing on refining entity linking and expanding the knowledge graph to capture emerging trends.

The cultural heritage domain in general and silk textiles, in particular, are characterized by large, rich and heterogeneous data sets. Silk heritage vocabulary comes from multiple sources that have been mixed up across time and space. This has led to the use of different terminology in specialized organizations in order to describe their artefacts. This makes data interoperability between independent catalogues very difficult. To address these issues, SILKNOW created a multilingual thesaurus related to silk textiles. It was carried out by experts in textile terminology and art historians and computationally implemented by experts in text mining, multi-/cross-linguality and semantic extraction from text. This paper presents the rationale behind the realization of this thesaurus.

This paper presents an Artificial Intelligence approach to mining context and emotions related to olfactory cultural heritage narratives, particularly to fairy tales. We provide an overview of the role of smell and emotions in literature, as well as highlight the importance of olfactory experience and emotions from psychology and linguistic perspectives. We introduce a methodology for extracting smells and emotions from text, as well as demonstrate the context-based visualizations related to smells and emotions implemented in a novel smell tracker tool. The evaluation is performed using a collection of fairy tales from Grimm and Andersen. We find out that fairy tales often connect smell with the emotional charge of situations. The experimental results show that we can detect smells and emotions in fairy tales with an F1 score of 91.62 and 79.2, respectively.

The global health situation due to the SARS-COV-2 pandemic motivated an unprecedented contribution of science and technology from companies and communities all over the world to fight COVID-19. In this paper, we present the impactful role of text mining and data analytics, exposed publicly through IRCAI's Coronavirus Watch portal. We will discuss the available technology and methodology, as well as the ongoing research based on the collected data.

We present a benchmark in six European languages containing manually annotated information about olfactory situations and events following a FrameNet-like approach. The documents selection covers ten domains of interest to cultural historians in the olfactory domain and includes texts published between 1620 to 1920, allowing a diachronic analysis of smell descriptions. With this work, we aim to foster the development of olfactory information extraction approaches as well as the analysis of changes in smell descriptions over time.

The MWE lexicon was extracted from the Gigafida 2.1 Corpus of Written Standard Slovene (https://www.clarin.si/noske/run.cgi/corp_info?corpname=gfida21) using specialized scripts for extracting data from corpora containing syntactic dependency annotations. The lexicon contains 5,242 Multiword Expressions with 12,358 examples from Gigafida 2.1. Each MWE entry (or sense) contains at least one and up to three extracted examples. MWEs were analysed using the JOS dependency parser system (http://nl.ijs.si/jos/bib/jos-skladnja-navodila.pdf) and were assigned matching syntactic structure IDs. The corpus sentences containing the MWE components and matching syntactic structure features were identified in the corpus and assigned to the corresponding headword or sense. MWEs variants (or variant senses) are linked with the senseKey attribute values, forming a MWE cluster of related variants or variant senses. A sample of MWE headwords also contains manually created sense division with descriptions of meaning for each sense.

This article presents the current outcomes of the MARCELL CEF Telecom project aiming to collect and deeply annotate a large comparable corpus of legal documents. The MARCELL corpus includes 7 monolingual sub-corpora (Bulgarian, Croatian, Hungarian, Polish, Romanian, Slovak and Slovenian) containing the total body of respective national legislative documents. These sub-corpora are automatically sentence split, tokenized, lemmatized and morphologically and syntactically annotated. The monolingual sub-corpora are complemented by a thematically related parallel corpus (Croatian-English). The metadata and the annotations are uniformly provided for each language specific sub-corpus. Besides the standard morphosyntactic analysis plus named entity and dependency and/or noun phrase annotation, the corpus is enriched with the IATE and EuroVoc labels. The file format is CoNLL-U Plus Format, containing the ten columns specific to the CoNLL-U format and four extra columns specific to our corpora. The MARCELL corpora represent a rich and valuable source for further studies and developments in machine learning, cross-lingual terminological data extraction and classification.

This paper presents recent developments and the content of the ssj500k training corpus, the largest and most widely used open-source collection of training data for Slovene language processing, which has been manually annotated with respect to segmentation, tokenisation, lemmatisation, JOS morphosyntax and dependency syntax, Universal Dependencies, semantic role labelling, named entities and verbal multi-word expressions. After a short history of the development of the corpus, we give an overview of the dataset as a whole, and the details of each annotation layer, including a survey of existing natural language processing tools that used it for training. Most ssj500k annotations were carried out using the dedicated Q-CAT querying-supported corpus annotation tool, which is also presented, and the directions for future development of the corpus are discussed.

Semantic annotation is the task of augmenting an unstructured textual document with semantic information, such as concepts from an ontology. In wikification, the Wikipedia is used as an ontology and its pages (articles) are regarded as (representations of) concepts. We describe an efficient approach for annotating a document with relevant concepts from the Wikipedia. A global disambiguation method based on constructing a mention-concept graph and computing pagerank over it is used to identify a coherent set of relevant concepts considering the input document as a whole. The presented approach is suitable for parallel processing and can support any language for which a sufficiently large Wikipedia is available. Several heuristics involved in the disambiguation of candidate annotations are discussed and an experimental evaluation of their influence is presented.

We describe an efficient approach for annotating a document with relevant concepts from the Wikipedia. A pagerank-based method is used to identify a coherent set of relevant concepts considering the input document as a whole. The proposed approach is suitable for parallel processing and can support any language for which a sufficiently large Wikipedia is available.

Data collected from mobile phones can be used to uncover underlying social network dynamics and individual’s behavioral patterns. Based on a Call Details Records dataset, we build a weighted, directed network and analyze it’s properties. In addition to node-level network measures we extract an extensive consumption and mobility-based feature set. We show that extracted network and consumption features can be used to model individual’s risk profile.

In this demo we describe a system called Event Registry (http://eventregistry.org) that can identify world events from news articles. Events can be detected in different languages. For each event, the system can extract core event information and store it in a structured form that allows advanced search options. Numerous visualizations are provided for visualizing search results.

QMiner is an open source analytics platform for performing large scale data analysis written in C++ and exposed via a Javascript API. The paper presents five main design elements which focus on focus on storage, online and real-time processing as well as fast prototyping and give QMiner unique advantages as a data analytics platform for processing streams of structured and unstructured data. These design elements are incorporated in a five layer architecture represented by 1) storage and indexing, 2) stream aggregate, 3) feature extractor, 4) linear algebra and 5) analytics layers. The functionality of the platform is demonstrated by three representative usage examples containing code fragments: text classification, time series prediction and community detection in graphs.

Event Registry is a system that can analyze news articles and identify in them mentioned world events. The system is able to identify groups of articles that describe the same event. It can identify groups of articles in different languages that describe the same event and represent them as a single event. From articles in each event it can then extract event's core information, such as event location, date, who is involved and what is it about. Extracted information is stored in a database. A user interface is available that allows users to search for events using extensive search options, to visualize and aggregate the search results, to inspect individual events and to identify related events.