Annotations with EARMARK in practice: a fairy tale

tags

Standoff

Notes

we need a way to write and store overlapping annotations

NOTER_PAGE: (1 . 0.32669789227166274)

adding annotations to TEI documents

NOTER_PAGE: (1 . 0.38065716057036575)

how to use EARMARK to link textual ranges of such documents to resources in the Linked Open Data.

NOTER_PAGE: (1 . 0.4038800705467372)

we need a way to refer to external entities, conceptualizations that live outside of a document and that are independent of it

NOTER_PAGE: (1 . 0.4637002341920375)

Literary texts are not isolated conceptual units: they live in a complex pattern of references

NOTER_PAGE: (1 . 0.5989461358313818)

multiple annotations need to be allowed by multiple encoders

NOTER_PAGE: (1 . 0.6762295081967213)

users must be allowed even to annotate documents for which they do not have write permissions

NOTER_PAGE: (1 . 0.6879391100702577)

Anno- tated text fragments, entities and annotators are identified by URIs in EARMARK

NOTER_PAGE: (1 . 0.7851288056206089)

EARMARK is based on existing Semantic Web technologies

NOTER_PAGE: (1 . 0.8220140515222483)

On the one hand, the element body contains two different speeches

NOTER_PAGE: (2 . 0.07084309133489461)

On the other hand, the element body also contains the stanzas and the lines the text is organzied in

NOTER_PAGE: (2 . 0.11124121779859485)

annotations are placed in a separate document with references to the parts of the docu- ment they refer to (standoff).

NOTER_PAGE: (2 . 0.6413415468856947)

techniques in which the annotations are em- bedded have the advantage of keeping all the information in a single coherent structure within a single file.

NOTER_PAGE: (2 . 0.7189695550351288)

Over- lapping markup is a complex topic without an universally accepted solution

NOTER_PAGE: (2 . 0.7810304449648712)

require the permissions for the annota- tor to modify the original document

NOTER_PAGE: (2 . 0.819672131147541)

the author may want to publish a document but not carry the burden of publishing someone else’s annotations as well

NOTER_PAGE: (3 . 0.07552693208430913)

need to address the data from the source original document; annotations must have a way to identify with precision the content they are referring to.

NOTER_PAGE: (3 . 0.13700234192037472)

Serious issues arise when the document is modified

NOTER_PAGE: (3 . 0.24414519906323184)

The Extremely Annotational RDF Markup, or EARMARK [9, 19], is an OWL 2 DL ontology that defines document meta-markup.

NOTER_PAGE: (3 . 0.6416861826697893)

hrough appropriate OWL and SWRL characterizations it can define structures such as trees or graphs and can be used to generate validity constraints

NOTER_PAGE: (3 . 0.7072599531615925)

make sure that annotations still refer to the correct piece of document once the document is changed.

NOTER_PAGE: (5 . 0.11592505854800937)

authors can be annotators themselves, allowing the use of different techniques that may reduce the ability of others to anno- tate the same documents.

NOTER_PAGE: (5 . 0.13992974238875877)

avoid some of the issues related to document modifications through the use of XPath pointer ranges.

NOTER_PAGE: (5 . 0.7101873536299765)

specify the par- ticular node of the whole docuverse to use as the context relative to which the offsets are expressed.

NOTER_PAGE: (5 . 0.7634660421545667)

create permanent and live connections between the inclusion and the original source of content

NOTER_PAGE: (7 . 0.07494145199063232)

This is a form of transclusion, similar to what has origi- nally been proposed by Ted Nelson for the Xanadu project

NOTER_PAGE: (7 . 0.11065573770491803)

The goal of the Xanadu project was to build a global workspace where all users could freely reuse, comment on and link to any piece of content.

NOTER_PAGE: (7 . 0.16334894613583137)

milestones (i.e. overlapping structures are expressed through a pair of empty elements to mark the boundaries of the “content”)

NOTER_PAGE: (7 . 0.5286885245901639)

ragmentation (i.e. overlapping struc- tures are split into individual, non-overlapping elements that are linked through id-idref pairs)

NOTER_PAGE: (7 . 0.5515222482435597)