Systems and methods for automatically identifying and linking names in digital resources

A method for identifying relationships among digital resources, the method comprising:

• searching within parsed data of a first digital resource for occurrences of name-like-strings, comprising at least one of the following steps:
• searching within the parsed data for specialized formatting,
• searching within the parsed data for a string satisfying a regular expression,
• analyzing the parsed data using a grammar constructed for a class of name or digital resource, performing a text search within the parsed data against a reference list of names, or
• performing a text search within the parsed data to identify a string having at least one of a known prefix or suffix used in a terminology,
• for each name-like-string identified in the parsed data:
• creating metadata marking an exact location in the parsed data of the occurrence of said name-like-string,
• obtaining a first unique content identifier by one of:
• matching said name-like-string against the names or labels of entities,
• wherein said entities are represented as information objects stored in a repository, and wherein each information object is identified by a persistent globally unique identifier, or
• creating a new unique content identifier for the name-like-string conforming to a grammar of a class of name or resource,
• creating a new relationship associating said first unique content identifier with the first digital resource, wherein said relationship comprises said first unique content identifier for said name-like-string, a second unique content identifier for said first digital resource, and metadata about said association, and
• storing said relationship in a database,
• wherein said first and second digital resources may be identified by at least one unique content identifier comprising a Digital Object Identifier (DOI), a Handle, an Archival Resource Key (ARK), a Persistent Uniform Resource Locator (PURL), a Universal Unique Identifier (UUID), a Globally Unique Identifier (GUID), a Uniform Resource Identifier (URI), or a database accession,
• wherein each information object identifier is one of a Digital Object Identifier (DOI), a Handle, an Archival Resource Key (ARK), a Persistent Uniform Resource Locator (PURL), a Universal Unique Identifier (UUID), a Globally Unique Identifier (GUID), or a Life Sciences Identifier (LSID),
• wherein said metadata about the association comprises at least one of said name-like-string, said exact location of said name-like-string in said parsed data, a number of occurrences of said name-like-string in the first digital resource, and a class of entity or resource identified by said first content unique identifier,
• inferring a plurality of previously unrecognized relations among digital resources, the method comprising:
• building a graph of information objects, digital resources and relations among information objects and digital resources retrieved from a database, wherein
• each information object comprises structured text, data, and metadata representing an entity in digital form, said entity being one of:
• name, taxon, exemplar, practitioner, or nomos,
• each information object is identified by at least one persistent globally unique identifier, wherein said persistent globally unique identifier resolves to a service or a second digital resource encoding said structured text, data, and metadata for said information object,
• said graph comprises a set of information objects and relations among said information objects, each relation linking a pair of information objects,
• each said pair comprising entities of type:
• name-name, name-taxon, taxon-taxon, taxon-exemplar, exemplar-exemplar, nomos-taxon, practitioner-name, practitioner-taxon, practitioner-exemplar, or practitioner-nomos, and
• each said pair including a relation representing one of:
• synonymy, homonymy, orthographic variation, equivalence, membership, supertype, or subtype,
• new relations among said information objects are inferred based on a set of encoded rules, and
• the information objects in said graph may be traversed over said relations;
• said graph further comprising a set of digital resources and relations among said information objects, each relation linking a pair of information object and digital resource,
• each said pair comprising one of a digital resource and one of an information object,
• said graph further comprising a set of digital resources and relations among said digital resources, each relation linking a pair of digital resources,
• analyzing said graph, wherein the analysis
• applies a metric to the relations among information objects and digital resources,
• traverses said graph from a third unique content identifier for a digital resource over said relations to identify a plurality of information objects and digital resources related to said third unique content identifier,
• retrieves plurality of related unique content identifiers from among the digital resources comprising said graph, and
• enumerates said plurality of related unique content identifiers.