Smith's Dictionary of Greek and Roman Biography and Mythology
The Perseus Digital Library’s digital edition of William Smith’s A Dictionary of Greek and Roman Biography and Mythology is based off of an 1873 printing by John Murray and includes all 3 volumes of this monumental reference work in a single TEI-XML file. This important reference work can be browsed or searched online as part of the Perseus Digital Library and has been featured there since 2004. The dictionary itself contains tens of thousands of entries for classical named entities according to the principle laid out by the original editors: “the names of all persons are inserted, who are mentioned in more than one passage of an ancient writer.” Consequently, Smith’s remains one of the most comprehensive if somewhat dated public domain sources for obtaining biographical and prosopographical information for named individuals from the ancient world.
The biographical entries vary greatly in length but each one does have a minimal level of additional tagging beyond basic structural encoding. All Greek words have been encoded within the entries and all dates and bibliographic citations have been automatically tagged, with varying levels of accuracy. Individual TEI-XML entries can be downloaded for each biographical article by clicking on the XML link at the bottom of each entry page (as for the Herodotus example below). Alternatively the entire XML file for Smith’s Dictionary can also be obtained from Github.
Ultimately, the Perseus Digital Library hopes to enhance the Smith’s Dictionary through an improved and more accurate citation encoding process, to make those citations machine actionable and referenceable to classical texts online, and to provide fully compliant linked data URIs for every entity within the dictionary. The Perseus Digital Library is in the midst of a move towards supporting fully compliant linked data by assigning stable identifiers to every primary and secondary source text and data set in its repository. There are two main classes of stable identifiers for the Smith’s data: identifiers for the personal name entities that are described by the dictionary entries, and identifiers for the dictionary entries themselves. So the entry for the historian Herodotus, for example, should have a citable stable URI for the name Herodotus, and a separate citable stable URI for the entry for this name in Smith’s Dictionary
Perseus Stable URIs were first assigned to each personal name entity within the dictionary. So, for example, the name entity for the historian Herodotus has the citable stable URI: http://data.perseus.org/people/smith:herodotus-1.
In addition, the Perseus Digital Library is using Canonical Text Service URNs for all primary sources and while it’s still not fully clear if CTS URNs are the best approach for secondary sources like the Smith’s Dictionary, it is an ongoing experiment. The Smith’s Dictionary was the first secondary source to which URNs were assigned to and it was then made available in a CTS repository. It has the URN urn:cts:pdlrefwk:viaf88890045.003.perseus-eng1. The citation scheme for this text has 2 levels: alphabetic entry and person entry. So, the full stable URI for the Herodotus entry in the text is: http://data.perseus.org/citations/urn:cts:pdlrefwk:viaf88890045.003.perseus-eng1:H.herodotus_1.
The Perseus Project has made this distinction between the named entity identifiers and the text entry identifiers so that references in a linked data environment can be semantically meaningful and can support both linking to entries within Smith and the reuse of the data within other digital classics projects that make use of named entity and/or biographical data.
An example of how this has been applied can be found in the Perseids Journey of the Hero project. Students were annotating entries in the Smith’s dictionary -- linking them to places, bibliographic references, attestations, and so forth. One aspect of this work has been to annotate the social relationships between the people mentioned in the biography entries. When an annotation is targeting a person name, the target of that annotation uses the URI for the named entity. But when an annotation is targeting explanatory text within the entry itself (such as the description of the relationship, or a mention of a place) the annotation targets the CTS URN reference for the text of the entry.
As the Perseus and Perseids projects continue their research into linked data, CTS URNs, and the potential data reuses of secondary sources such as the Smith’s dictionary, further applications and experiments will likely be developed.