Despite advancements in NLP-based mental health screening, research remains predominantly English-centric, leaving under-resourced languages insufficiently explored. This study investigates depression detection in Modern Greek social media through a series of experiments. We benchmark traditional machine learning (ML) models against transformer architectures (GreekBERT, GreekSocialBERT, mBERT, and XLM-R) under two settings: a topic-oriented control corpus and a high-similarity stress-test contrasting a gold case of a depressed user with a matched control. Transformer models consistently outperform ML models (F1 = 0.95) but offer limited interpretability. To address this limitation, we incorporate LIWC-derived psycholinguistic features with SHAP explanations to examine model behavior in relation to established linguistic markers. The analysis reveals linguistic patterns consistent with depressive symptoms, such as reduced work-related engagement, social withdrawal, and the motivational deficits characteristically linked to anhedonia in clinical literature. Overall, the results provide a baseline for depression detection in Modern Greek and underscore the importance of grounding automated screening in clinically interpretable evidence.

This paper presents the first automatic speech recognition (ASR) system for Cypriot Greek, a non-standardized variety of Modern Greek with distinctive phonological, lexical, and orthographic characteristics. We adapt Whisper, a state-of-the-art multilingual ASR model, to Cypriot Greek through fine-tuning on the mozilla common voice spontaneous speech dataset for Cypriot Greek. The phonological and lexical divergence of Cypriot Greek from Standard Modern Greek poses significant challenges for mainstream ASR, particularly under conditions of limited training data and dialectal variation. Results demonstrate that whisper-medium achieved a best word error rate (WER) of 37.85%, while whisper-large-v3 consistently outperformed it, reaching a minimum WER of 33.93%. In the light of these findings, increased model size, combined with targeted fine-tuning on normalized dialectical data, significantly improves recognition accuracy, indicating that careful handling of orthographic and dialectical variation provides an effective path for ASR adaptation to low-resource varieties.

Pontic Greek is a living but endangered Modern Greek dialect that lacks publicly available AI-oriented speech resources and ASR benchmarks. This work reports on the first systematic inference-only (zero-shot) ASR evaluation on authentic Pontic speech. Progress on Pontic ASR is hindered by two coupled challenges: the scarcity of transcribed speech data and the absence of a standardized orthography, which makes it difficult to create consistent reference transcriptions for evaluation. We address these challenges by releasing a new speech corpus of contemporary Pontic as spoken in Northern Greece, derived from natural conversations and provided with manual, utterance-level, time-aligned transcriptions. To reduce annotator bias and increase practical usability, we collect community evidence on written-form preferences via a small questionnaire and use the observed patterns to guide a consistent Greek-script transcription scheme. We use this corpus to perform inference-only (zero-shot) ASR evaluation, benchmarking four state-of-the-art pretrained speech recognition models under a unified evaluation protocol. Results show that zero-shot recognition remains challenging, establishing baseline figures and underscoring the need for dialect-specific data and adaptation.

We report on the development of the first treebank and parser for Eastern Cretan in the framework of Universal Dependencies (UD). Eastern Cretan is a living but under-resourced dialect of Modern Greek. We have worked on the transcription of oral material and relied on active annotation and knowledge transfer from GUD, a treebank of Standard Modern Greek. Along with its other phonological and morphosyntactic differences from Standard Modern Greek, Eastern Cretan (and other varieties of Modern Greek) makes heavy use of euphonics and voicing that have not been included in the UD annotation guidelines so far. We have provided annotation guidelines for East Cretan euphonics and voicing and included them in the models. Knowledge transfer from the treebank of Standard Modern Greek to the dialectal models helped to initiate annotation via an active annotation procedure

UD_Greek-GUD (GUD) is the most recent Universal Dependencies (UD) treebank for Standard Modern Greek (SMG) and the first SMG UD treebank to annotate Verbal Multiword Expressions (VMWEs). GUD contains material from fiction texts and various sites that use colloquial SMG. We describe the special annotation decisions we implemented with GUD, the pipeline we developed to facilitate the active annotation of new material, and we report on the method we designed to evaluate the performance of models trained on GUD as regards VMWE identification tasks.

This paper presents a methodological approach for establishing control corpora in the context of depression detection in the Modern Greek language. We discuss various methods used to create control corpora, focusing on the challenge of selecting representative samples from the general population when the target reference is the depressed population. Our approach includes traditional random selection among Twitter users, as well as an innovative method for creating topic-oriented control corpora. Through this study, we provide insights into the development of control corpora, offering valuable considerations for researchers working on similar projects in linguistic analysis and mental health studies. In addition, we identify several dominant topics in the depressed population such as religion, sentiments, health and digestion, which seem to align with findings consistently reported in the literature

We introduce a new corpus, named AIKIA, for Offensive Language Detection (OLD) in Modern Greek (EL). EL is a less-resourced language regarding OLD. AIKIA offers free access to annotated data leveraged from EL Twitter and fiction texts using the lexicon of offensive terms, ERIS, that originates from HurtLex. AIKIA has been annotated for offensive values with the Best Worst Scaling (BWS) method, which is designed to avoid problems of categorical and scalar annotation methods. BWS assigns continuous offensive scores in the form of floating point numbers instead of binary arithmetical or categorical values. AIKIA’s performance in OLD was tested by fine-tuning a variety of pre-trained language models in a binary classification task. Experimentation with a number of thresholds showed that the best mapping of the continuous values to binary labels should occur at the range [0.5-0.6] of BWS values and that the pre-trained models on EL data achieved the highest Macro-F1 scores. Greek-Media-BERT outperformed all models with a threshold of 0.6 by obtaining a Macro-F1 score of 0.92

Pomak is an endangered oral Slavic language of Thrace/Greece. We present a short description of its interesting morphological and syntactic features in the UD framework. Because the morphological annotation of the treebank takes advantage of existing resources, it requires a different methodological approach from the one adopted for syntactic annotation that has started from scratch. It also requires the option of obtaining morphological predictions/evaluation separately from the syntactic ones with state-of-the-art NLP tools. Active annotation is applied in various settings in order to identify the best model that would facilitate the ongoing syntactic annotation.

We present a cleansed version of the multilingual lexicon HURTLEX-(EL) comprising 737 offensive words of Modern Greek. We worked bottom-up in two annotation rounds and developed detailed guidelines by cross-classifying words on three dimensions: context, reference, and thematic domain. Our classification reveals a wider spectrum of thematic domains concerning the study of offensive language than previously thought Efthymiou et al. (2014) and reveals social and cultural aspects that are not included in the HURTLEX categories.

The project XXXX is developing a platform to enable researchers of living languages to easily create and make available state-of-the-art spoken and textual annotated resources. As a case study we use Greek and Pomak, the latter being an endangered oral Slavic language of the Balkans (including Thrace/Greece). The linguistic documentation of Pomak is an ongoing work by an interdisciplinary team in close cooperation with the Pomak community of Greece. We describe our experience in the development of a Latin-based orthography and morphologically annotated text corpora of Pomak with state-of-the-art NLP technology. These resources will be made openly available on the XXXX site and the gold annotated corpora of Pomak will be made available on the Universal Dependencies treebank repository.

We evaluate manually five lexical association measurements as regards the discovery of Modern Greek verb multiword expressions with two or more lexicalised components usingmwetoolkit3 (Ramisch et al., 2010). We use Twitter corpora and compare our findings with previous work on fiction corpora. The results of LL, MLE and T-score were found to overlap significantly in both the fiction and the Twitter corpora, while the results of PMI and Dice do not. We find that MWEs with two lexicalised components are more frequent in Twitter than in fiction corpora and that lean syntactic patterns help retrieve them more efficiently than richer ones. Our work (i) supports the enrichment of the lexicographical database for Modern Greek MWEs’ IDION’ (Markantonatou et al., 2019) and (ii) highlights aspects of the usage of five association measurements on specific text genres for best MWE discovery results.