Orazio O.

Key Skills

Software

Other

Top Subject Matter

Authorship attribution / AI-text detection for Italian

Corpus linguistics and NLP pipeline evaluation for Italian

NLP and annotation for language technologies

Top Data Types

Text

Top Task Types

Fine-tuning

Freelancer Overview

Distinguishing AI-generated from human-written Italian text (2026). Core strengths include Other. Education includes Master of Science, University of Turin (2025) and Bachelor of Arts, University of Turin (2025). AI-training focus includes data types such as Text and labeling workflows including Evaluation, Rating, and Fine-tuning.

Labeling Experience

MSc Language Technologies and Digital Humanities (2025–Present), coursework on NLP and annotation

OtherTextFine-tuning

Completed NLP-focused coursework involving computational linguistics and annotation-related components, supporting supervised NLP training workflows. Applied methods and tools for NLP pipelines, including annotation practice and evaluation concepts as part of ML and deep learning study. The training emphasized using annotated linguistic structures as inputs to modeling and iterative experimentation. • Coursework: Computational Linguistics covering ML, deep learning, LLMs, and NLP pipelines with annotation • Coursework: Languages, Methods and Tools for NLP (focus on NLP methods/tools) • Annotation-related instruction and integration into model-oriented pipeline thinking • Evaluation of approaches in the context of language technologies

2025 - Present

Collocational shift in Italian web discourse before and during COVID-19 (2026)

OtherText

Conducted corpus-based analysis of collocational shifts using web corpora to characterize linguistic constructions before and during COVID-19. Compared patterns between pre-pandemic and pandemic datasets and interpreted constructional polysemy revealed through concordance and word sketch outputs. Although not framed as manual annotation, the study produced labeled linguistic evidence used for downstream interpretation and feature-level evaluation. • Comparative analysis of Italian virale collocations in itTenTen16 vs itTenTen20 • SketchEngine word sketches and CQL-based querying • Concordance analysis to quantify changes in constructions and frequency • Tool-assisted corpus linguistics workflows in support of labeled linguistic findings

2026 - 2026

Distinguishing AI-generated from human-written Italian text (2026)

OtherText

Built an AI-generated-vs-human classification pipeline by extracting linguistic features from Italian texts. Trained a logistic regression model using cross-validation and performed feature ablation to identify discriminative signals. The work focused on generating training/evaluation data representations from raw text inputs. • Dataset of 200+ Italian texts across 10 topics • Feature extraction using spaCy (lexical diversity, POS distribution, dependency depth, morphological richness) • Logistic regression with 10-fold cross-validation • Feature ablation comparisons for signal selection

2026 - 2026

Education

Work History