Generalized and Integrated Multidimensional Referencing

Relation References
Relations are key logical constructors based on Subject and Object references, and which typically define and qualify the relation between their subject and object references.

General Referencing
Whether qualified as parts of relations or through general (e.g. unqualified) references, references can take different types. These reference types match natural dimensions and should probably be called reference dimensions, mapping identity space, time/period, physical space, as well as electronic space references. The most important and commonly used include:

• Identification-based: Identity referencing (e.g. Primary keys, GUIDs, RFIDs, Model/Series, etc.)
• Electronic Locations: Electronic (e.g. Web) document and asset referencing (e.g. URIs, Queries, email addresses, etc.)
• Period-based: Single or repeating (e.g. pattern/rule -based) Period, Date, Time, Interval, Duration referencing (e.g. ISO 8601)
• Space-based: Secific or vague (e.g. Country, or State, or City, or PostalCode) spacial and physical references (e.g. addresses, geo-location, phone or fax number, airport code, etc.)
• Rule-based: Rule and query based references (e.g. xpath, xquery, sql, or other query types)

Genralized and Integrated Reference Support
Whether internally, inside a document, outward to an external document, or externally between documents and resources, relations typically need to reference resources through any and all reference type.

Genralized and Integrated Relationship Management Support
With an integrated and generalized referencing approach, relations can be generalized and integrated also, in addition to being better articulated.

Standards-based Referencing
Any effective integrated and generalized referencing approach needs to be standards-based and focus mostly on integrating existing standards for more generalized referencing

String-based References
As referencing standards like ISO8601, as well as URL/URI/URN specifications have demonstrated, the most practical and effective representation for references is through character strings. Any reference integration and generalisation approach should follow the same path. It may also be emphasized that string-based references can easily be embedded anywhere, including in XML-tpye attribute nodes.

Reference Parsing / Writing API
While string references are most practical as well as both human and machine readable, it is often useful to parse and expand these reference strings into more structured elements and items. Similarily, often, more structured elements and items representing references, also need to be written back to strings. Accordingly, standards API for parsing and writing reference strings should be provided. Most API functions should also be parametrized to provide optimal flexibility and compatibility.

StratML Relations
Relations are key to StratML documents, sections, and collections. They are also key manageable and expandable document structure. More so, they are key to the value of strategy and performance plans and report, as well as to the relevance and value of the organizations that require these plans and reports. Consequently, it is crucial that StratML Part3 adopts a structured approach to relations, as proposed in previous sections, including an integrated and generalized approach to referencing.

Options
Relations are key to StratML documents, sections, and collections. They are also key manageable and expandable document structure. Accordingly, the following sections will briefly present such an integrated and generalized approach, as it has been implemented in the internal representation used to manage the current StratML portal. Other approaches to structured referencing integration and generalization are possible and the StratML portal team will be happy to contribute, but an effective approach is required. One is introduced below, it has been tested and implemented. The introduction focuses on the overall structure and does not discuss or fully justify the implied choices.

Single Instances of Unidimensional References
The simplests and possibly most common form of references, are single instances of unidimensional references. These typically refer to a single resource, from a single reference dimension. Common examples may include: a document URL, a phone number, a mailing address, a GUID, a date or date-time.

Iterated Instances of Unidimensional References
A reference can also refer to a sequence of resources in a single referencing dimension. Examples may include: a sequence of document URLs, a sequence of phone numbers, a sequence of mailing addresses, a sequence of GUIDs, a sequence of dates or date-time references.

Single Instances of Multidimensional References
References can also refer to a resource from multiple dimensions, combining different types of references to a single resource. Examples may include: a person SSN, address, phone, fax, sex, date of birth, and date-time of last meeting (e.g. modification date), or a record's database URL, access string (ex: userID, password), table reference, and primary key.

Iterated Instances of Multidimensional References
Many times, and probably growingly so, a reference may refer to a sequence of resources, each from a variety of dimensions. Examples may include: a family's sequence of personal SSNs, addresses, phones, faxes, sexes, dates of birth, and date-time of last meetings (e.g. modification dates), or a sequence of records' database URLs, access strings (ex: userIDs, passwords), table references, and primary keyes.

Rule-based Referencing Generalization and Integration
Another type of common references are rule-based queries that can refer to single or sequence of resources, possibly from any referencing dimension or dimension combination. Typical examples may include: SQL, XQuery, XPath or other types of queries. Not only can these be used alone, but they can also typically be combined with other forms of references, including, for example, a database URL, an access string (e.g. userID, Password), and a query. More so, even simple single instance unidimensional references, like a phone number, can be interpreted as a simple query to match the number provided. Consequently, to better generalize references, all reference strings should be interpreted as rule-based queries. As there are a few standards-based query languages, at least XPath, XQuery, and SQL should be supported, along with the default "plain" referencing scheme.

Generalized Integrated Rule-based Referencing Schema Standards
Generalized Integrated Rule-based Referencing Schema standards are not yet available, but a working Schema and API already exist and are available for standardisation, especially as they are based on and already support most currently existing referencing standards. These specifications are not fully included here, but are available for standardization and upon formal requests.