Talk:NISO-I2

Major points from email-based discussions

1. Herbert introduced I2 midterm report and initiated the discussion -- Date: Mon, 12 Jul 2010

I have only briefly glanced at the mid-term report, but I found no indication that this work is informed by a Linked Data perspective. Nor did I find immediate indications that it could not be implemented in a way that aligns with Linked Data.

Most of the use cases listed in the report involve libraries and hence seem relevant to this group.

2. Jeff's comments and proposed I2 OWL ontology

HTTP URIs are capable of identifying anything imaginable.

Here are some words others can throw rocks at.

I^2 identifiers should be http URIs so that agents anywhere in the world can easily dereference the identifier and get back information about the institution. This http identifier should behave according to Linked Data conventions so that humans and machines can use HTTP content negotiation to retrieve a convenient representation (HTML, XML, RDF, etc.) without confusing the institution’s identity with the identity of the various web document representations.

http://www.w3.org/DesignIssues/LinkedData.html http://www.w3.org/TR/cooluris/

See my earlier OWL and example mockup for details.

[this part is attached to the comments]

As a more realistic proposal, I converted the NISO I2 metadata element set to OWL (i2.owl) and mocked up some RDF for a sample institution that could be stored in their registry (about.rdf). A UML class diagram for the OWL is also attached. Here’s how I suggest the institution identifier behave according to Linked Data:

http://i2.niso.org/institution/1 (303 redirect to...)

http://i2.niso.org/institution/1/ (content-negotiate to…)

http://i2.niso.org/institution/1/default.html (A human-readable representation)

http://i2.niso.org/institution/1/about.rdf (see the attached about.rdf)

I used a sequential number as the opaque institution ID and guessed on the http://i2.niso.org/institution/ part. There are various details worth quibbling about.

3. Emma's question about bringing this to the teleconference --Date: Tue, 27 Jul 2010

Following up on your suggestion that a collective feedback could be provided by LLDXG, I wonder if you'd like to see the topic raised during our call on Thursday. This could be limited to a few minutes, just to quickly ascertain whether XG members roughly agree on the thrust of a response and to give interested members an action to liaise on the wording of a response for submission in time for the deadline.

The response could be signed "Members of the Library Linked Data Incubator Group" and there should be a sentence explaining that the issue came up on the list but that the comments represent informal feedback, not an official response from the group.

4. Marcia's suggestions of options -- Tuesday, July 27, 2010

Option 1. If Herbert has already a proposed piece of the comments, others can add their suggestions. (If he will not bring a proposed piece, it will be difficult to generate a collective comments from the teleconference since many of us may have not studied it carefully. )

Option 2. To discuss through listserv on these points before teleconference. If there are good discussions we then get further feedback on the teleconference.

2.1 Based on Herbert’s general feedback, this report had “no indication that this work is informed by a Linked Data perspective. Nor did I find immediate indications that it could not be implemented in a way that aligns with Linked Data.”[2] What we can do is to provide a general comment to request the I2 group to bring Linked Data aspect into its next phase of work.

2.2. The teleconference can determine if the purposes of I2 have covered enough in its ‘information supply chain’ if aligning it with linked data.

According to the midterm report, “The Institutional Identifier (I2) has two purposes:

1. Identify institutions of all types. Each identifier will be globally unique and will represent sufficient metadata to differentiate institutions unambiguously.

2. Identify institutions engaged in the selection, purchase, licensing, storage, description, management, and delivery of information (“information supply chain”).” [1]

2.3 I believe we can also discuss if we should request that URI be considered in the final version. Jeff’s comments [3] can be summarized to respond this point.

From the original document Q&A (emphasis added by MZ)

“Q: Section 2 (Environment and structure of the Institutional Identifier) and Section 3 (Features of the Institutional Identifier and potential identifier standards) both mention the concept that a business specific registry could add a URI to the central registry to point back to extended information about the institution that is stored in the business specific registry. Is this element represented in the proposed metadata?

A: The initial version of the metadata did not include the URI. This element will become part of the final version of the metadata if it is deemed a valuable addition to the standard. We expect in the final development phase to look very closely at the utility of I2 to support business-specific scenarios as outlined in Section 5, Scenarios Illustrating Use of the Institutional Identifier. These investigations, together with the feedback we receive from interested parties, should enable us to answer this question.” [1]

2.4 The report has another answer to this question. It is relevant to LLD too. Is there any thought?

From the original document Q&A

“Q: What about institutional identifiers in common use in libraries, such as the MARC Organization Codes and OCLC codes?

A: These codes pre-date the Web and are not “resolvable,” meaning that they do not resolve to a web-based registry of information about the institution. They are also functional for their intended purpose in identifying holdings belonging to a specific library for union catalogs and interlibrary lending, but they are not always globally unique and not extensible to other purposes that require links to variant identifiers, to related institutions and the ability to link forward to business specific registries.”[1]

[1] http://www.niso.org/workrooms/i2/midtermreport/I2midterm2010.pdf

[2] 7/12/10 10:56 AM, "Herbert Van de Sompel" <hvdsomp@gmail.com> email

[3] Jeff Young email

6 Tom's suggestion for handling the comments as 'members'-- Date Wednesday, July 28, 2010

Formally speaking, since we will not have time to ensure consensus on the points more systematically, I suggest the comments be submitted from "members of the LLD XG" with a cover note referencing any list and telecon discussion and with a disclaimer that, given the tight deadline, the comments have not been approved by the group as a whole.

7. Background: Original call for comments from NISO

From: "Cynthia Hodgson" <chodgson@niso.org> Date: July 12, 2010 4:28:11 PM CEDT To: "Cynthia Hodgson" <chodgson@niso.org> Subject: [Digipres] NISO Seeking Feedback on Institutional Identifier Midterm Report -- responses needed by August 2 Reply-To: <chodgson@niso.org>

The NISO Institutional Identifier (I2) Working Group (WG) has released a midterm report: http://www.niso.org/workrooms/i2/midtermreport/

The NISO I2 WG is soliciting feedback on the report and guidance for the next steps in developing this standard from individuals and groups involved in the digital information transactions. Stakeholders include publishers/distributors, libraries, archives, museums, licensing agencies, standards bodies, and service providers, such as library workflow management system vendors and copyright clearance agencies. Anyone involved at any level in the distribution, licensing, sharing or management of information is invited to participate.

Please read the information below and participate in the evaluation of our midterm work by reading the midterm release document and answering a few questions about each development area. You are the stakeholders for this information standard. We must work to ensure that it meets your needs, so your input is very valuable and important to us.

BACKGROUND: NISO established the working group in 2008 to develop an institutional identifier (I2) to uniquely identify institutions engaged in the digital information workspace. The goal of the I2 Working Group is to develop an institutional identifier that is globally unique, robust, interoperable, scalable and able to integrate smoothly with current digital information workflows. The working group is currently at the midterm of its efforts and hopes to complete its draft specification by December, 2010. Community input was requested through surveys and conferences to refine the objectives, create the metadata and identify scenarios of need. We are currently soliciting midterm review to provide confirmation of our work to date, course correction as needed and to ensure that we have identified and are addressing all the issues surrounding this critical enabling standard.

THE PROBLEM SPACE: 

Obtaining, using, sharing, storing and managing information often involves multiple institutions across the digital information space. These institutions must be able to identify each other and to trust that the identification is both correct and unique. The information managed may itself be digital (e.g., the licensing of an e-book) or analog information that is managed over the digital information space (e.g., interlibrary loan of a physical book). Currently, there are many identifiers in use, ranging from simple naming to established codes. However, no single identifier that is globally unique, trustworthy, and able to capture relationships among institutions and variant legacy identifiers for institutions currently exists. As a result, transactions are locked into proprietary workflow silos and management of all the digital information activities of an institution are not integrated.

THE PROPOSED SOLUTION: The I2 is proposed as a globally unique, robust, scalable and interoperable identifier with the sole purpose of uniquely identifying institutions. The I2 consists of two parts: an identifier standard that includes the metadata needed to uniquely identify the organization -- including documenting relationships with other institutions that are critical for establishing identity -- and a framework for implementation and use.

The I2 is envisioned as a simple, core identifier with the sole purpose of identifying institutions in a robust and trustworthy manner. Workflow-specific implementations, such as regional ILL collaborations or ebook licensing services, will leverage the I2.

THE BENEFIT: Institutions will only have to request and reuse a single identifier. Institutions will be able to robustly identify every institution engaged in an information transaction. Institutions that engage in many different information transactions or that work with many different institutions will be able to track and manage institutional activities across multiple workflows through the use of a single, authoritative identifier.

The Midterm status report and review survey are available at the following link. Please respond by August 2, 2010.

http://www.niso.org/workrooms/i2/midtermreport/

Thank you very much for your support of this lynchpin digital information standard. Your input is very valuable to us and will be carefully studied and considered. Please download the report and keep it open to assist you in completing the survey.

[Note: This message has been cross-posted to obtain wide input.]

Cynthia Hodgson NISO Technical Editor Consultant National Information Standards Organization Email: chodgson@niso.org Phone: 301-654-2512