Open Collaborative Platform for Text Annotation, E-Lexicography, and Lexical Linking
Maia is an an open and collaborative web tool based on semantic web and linked open data technologies for text annotation, e-lexicography, and lexical linking, primarily designed and developed for use by digital humanists. It was conceived, designed, and developed to respond to the expressed need of scholars to work within the same environment, combining a reference corpus for resource construction and formal connections between lexical elements and textual elements.
The collaborative working environment developed encompasses both text and lexicon, allowing for the annotation of the text with arbitrary information and constructing the lexicon based on established lexical models. Through linking between text annotations and lexicon elements, it becomes possible to establish a formal connection between lexical entries, forms or senses, and their respective occurrences in a reference corpus, which can be structured into folders and subfolders.
Main features of MAIA
Maia provides an integrated environment that offers manifold advantages:
- within a task of constructing (or revising) a lexical or terminological resource, it is possible to consult the documents used as textual sources within the same environment. The ability to search through texts, annotate specific parts, and link these annotations to elements in the lexicon can, for instance, allow the definition of a polysemous term in its various meanings and connect the various senses created to their respective occurrences in the text. This way, it will be possible to have a series of contexts for each lexical entry, form, or sense defined in the lexicon, both for internal reference within the tool and for potential exports as examples of occurrences in the form of electronic dictionaries.
- the collaborative nature of Maia allows for the creation of shared work projects (whether they involve text annotation, lexicon construction, or a combination of both) involving multiple scholars, each of whom can focus on specific portions of the dataset, enabling parallel work.
Development prospects for Maia
- extension of the search functionalities, both for the corpus and the lexicon, with a dedicated section for linguistic-based search supported by the lexical resource.
- ability to create and edit ontologies within the environment: this work will allow for direct connections of text annotations to concepts within the ontology and linking senses from the lexical resource to these concepts. This will introduce a conceptual layer into the overall resource, distinct from the linguistic-semantic layer. In the first instance, Maia will allow importing and visualising OWL ontologies while the editing part will be developed later.
- data import and export features for the textual corpus and lexical resources. Initially, it will be possible to import pre-annotated texts in a format similar to CoNLL (which can then be manually reviewed in the system) and lexical resources in RDF format compliant with the OntoLex-Lemon model40. The data will, therefore, be exportable in the same formats. Subsequently, additional import and export formats will be included to accommodate the needs of users from various communities.
- upgrades involving the annotation component (to facilitate the annotation work in a collaborative context, TextO already includes versioning services).
- creation of links between annotations: through this feature, it will be possible to define, for example, anaphoric or syntactic relationships.
MAIA is under development.
For more details on the first version of MAIA:
Giovannetti, E., Albanesi, D., Bellandi, A., Marchi, S., Papini, M., & Sciolette, F. (2024). Maia: an Open Collaborative Platform for Text Annotation, E-Lexicography, and Lexical Linking. Umanistica Digitale, 8(18), 27–52. https://doi.org/10.6092/issn.2532-8816/19705