LiGre: A Lemma Bank for Ancient Greek via Linked Data
A new resource developed within the wider ANNOPHIS research context is now available: LiGre (Linking Greek), a linked open data lemma bank for Ancient Greek.
LiGre publishes Ancient Greek lemmas as structured and interlinked data, making lexical information easier to query, reuse, and connect with other linguistic resources. Its data model builds on established Linked Data standards and follows the approach developed by the LiLa Knowledge Base for Latin, extending this infrastructure to Ancient Greek.
The lemma bank can be searched online, while the underlying data are also accessible as RDF through a SPARQL endpoint and as individual Linked Data resources.
LiGre contributes to the broader ANNOPHIS goal of developing interoperable digital resources for the analysis and enrichment of historical-language data. In particular, it supports the project’s Linguistic Analysis use case by providing structured lexical data that can be linked to annotated texts and reused in computational workflows for linguistic analysis.
Explore LiGre: https://ligre-lod.github.io/
Managing and Accessing Text Collections
YouSayData has developed the ANNOPHIS Text Service, a lightweight service for managing and accessing structured text collections within the ANNOPHIS infrastructure.
The service exposes texts through a modern web API and supports fine-grained retrieval of textual passages using Canonical Text Services (CTS) identifiers. Researchers and applications can retrieve complete texts or specific passages, work with passage ranges, navigate between textual units, and access catalog information associated with a collection.
The Text Service works with CEX (CITE Exchange) data and provides a consistent way for other components of the infrastructure to request textual material. By separating text storage and retrieval from the interfaces used to annotate or visualize texts, the service helps create a modular architecture in which the same textual sources can be reused across different tools and workflows.
Within ANNOPHIS, the Text Service provides the underlying text-access layer for annotation, visualization, and machine-assisted analysis. It enables other components of the platform to work with precisely identified passages of historical texts while maintaining a consistent and reusable representation of the source material.
Explore the ANNOPHIS Text Service: https://github.com/GhentCDH/annophis-text-service
Creating and Managing Structured Annotations
The Ghent Centre for Digital Humanities (GhentCDH) has developed a reusable Annotation Editor for creating and managing structured annotations on textual sources.
The editor provides an interface for selecting passages of text and creating, editing, linking, and removing annotations. Annotation types and their properties can be configured for different research contexts, allowing the same underlying infrastructure to support a wide range of annotation tasks.
The software follows the W3C Web Annotation model, helping to ensure that annotations are represented in a standardized and interoperable way. Its modular design makes it possible to reuse the editor across different digital humanities applications and annotation workflows.
Within ANNOPHIS, the Annotation Editor contributes to the technical foundation for manual annotation and the validation and enrichment of historical texts. It supports researchers in creating structured annotations while keeping the underlying annotation model reusable and interoperable.
Explore the Annotation Editor: https://github.com/GhentCDH/annotation-editor
Connecting ANNOPHIS to Machine-Learning Workflows
The Ghent Centre for Digital Humanities (GhentCDH) has developed the ANNOPHIS ML Hub, a component that connects the ANNOPHIS infrastructure to machine-learning and natural-language-processing workflows.
The ML Hub acts as an intermediary between textual data and different annotation models. Text can be submitted to one or more configured annotators, for example for part-of-speech tagging, named-entity recognition, or other forms of automatic linguistic annotation.
Different annotators can be combined into processing pipelines, allowing the output of one step to serve as input for the next. The infrastructure uses the LAPPS Interchange Format (LIF), a JSON-LD format for exchanging linguistic annotations between NLP tools and services.
Within ANNOPHIS, the ML Hub provides the technical bridge between automated text analysis and the project’s annotation environment. It supports machine-assisted annotation workflows in which automatically generated annotations can be inspected, validated, and further enriched by researchers.
Explore the ANNOPHIS ML Hub: https://github.com/GhentCDH/annophis-mlhub
Visualizing Annotated Text
The Ghent Centre for Digital Humanities (GhentCDH) has developed Annotated Text, a reusable component for visualizing annotations on textual sources in web applications.
The component makes it possible to display different kinds of annotations directly alongside the text they describe, including linguistic annotations, structural information, and other forms of textual enrichment. It supports different text formats and can be extended through custom text and annotation adapters.
Annotated Text supports the W3C Web Annotation format as well as custom annotation structures, making it suitable for use across different digital humanities projects and annotation workflows. Its reusable design allows annotated texts to be integrated into research interfaces without tying the visualization layer to a single application.
Within ANNOPHIS, this component contributes to the presentation and exploration of enriched historical texts. It complements the Annotation Editor by providing a visualization layer through which researchers can inspect annotations and their relationship to the underlying text.
Explore Annotated Text: https://github.com/GhentCDH/annotated-text