XCRI Forum: Recent Posts

XCRI Forum: Recent Posts http://www.xcri.org/forum/ XCRI Forum: Recent Posts en Fri, 18 May 2012 04:02:45 +0000 https://profiles.google.com/matt.southall on "Data Definitions Example Problem?" http://www.xcri.org/forum/topic.php?id=197#post-819 Thu, 17 May 2012 11:59:19 +0000 https://profiles.google.com/matt.southall 819@http://www.xcri.org/forum/ Thanks Craig. Craig Hawker on "Data Definitions Example Problem?" http://www.xcri.org/forum/topic.php?id=197#post-818 Tue, 08 May 2012 08:29:47 +0000 Craig Hawker 818@http://www.xcri.org/forum/ Matt, I believe the examples within the Data Definitions document are to try and avoid queries around undeclared namespace prefixes. In the case of the UCAS Course Code, you should probably be looking to use the vocabularies that the Course Data Programme have put in place. These can be found in the same place as the Data Definitions document. In particular they have a <code>ucasCode</code> type which you should use. The following is a complete example which validates (well, the identifier does): <code><?xml version="1.0" encoding="UTF-8"?> <catalog xmlns="http://xcri.org/profiles/1.2/catalog" xmlns:credit="http://purl.org/net/cm" xmlns:dcterms="http://purl.org/dc/terms/" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:mlo="http://purl.org/net/mlo" xmlns:xhtml="http://www.w3.org/1999/xhtml" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xcriTerms="http://xcri.org/profiles/1.2/catalog/terms" xmlns:courseDataProgramme="http://xcri.co.uk" xsi:schemaLocation="http://xcri.org/profiles/1.2/catalog <a href="http://www.xcri.co.uk/bindings/xcri_cap_1_2.xsd" rel="nofollow">http://www.xcri.co.uk/bindings/xcri_cap_1_2.xsd</a> <a href="http://xcri.org/profiles/1.2/catalog/terms" rel="nofollow">http://xcri.org/profiles/1.2/catalog/terms</a> <a href="http://www.xcri.co.uk/bindings/xcri_cap_terms_1_2.xsd" rel="nofollow">http://www.xcri.co.uk/bindings/xcri_cap_terms_1_2.xsd</a> <a href="http://xcri.co.uk" rel="nofollow">http://xcri.co.uk</a> <a href="http://www.xcri.co.uk/bindings/courseDataProgramme.xsd" rel="nofollow">http://www.xcri.co.uk/bindings/courseDataProgramme.xsd</a>" generated="2011-03-24T09:32:50.827Z"> <dc:description>test catalog</dc:description> <provider> <dc:identifier>http://www3.myinstitution.ac.uk/</dc:identifier> <dc:title>test provider</dc:title> <mlo:url>http://www3.myinstitution.ac.uk/</mlo:url> <course> <dc:description xsi:type="xcriTerms:careerOutcome">hello world</dc:description> <dc:identifier>http://data.open.ac.uk/course/a174</dc:identifier> <dc:identifier xsi:type="courseDataProgramme:ucasCode">V101</dc:identifier> <dc:subject identifier="KC.3">KC.3</dc:subject> <dc:subject>Literature and Creative Writing</dc:subject> <dc:title>Start writing fiction</dc:title> <mlo:url>http://www3.open.ac.uk/study/undergraduate/course/a174.htm</mlo:url> <abstract>If you've ever considered writing a novel or short stories, this practical introduction to writing fiction will get you started, firing your creativity and imagination.</abstract> <mlo:assessment> <xhtml:div> <xhtml:p>You can choose whether to submit your tutor-marked assignments (TMAs) on paper or online through the eTMA system. You may want to use the eTMA system for some of your assignments but submit on paper for others. This is entirely your choice.</xhtml:p> <xhtml:p>The second TMA builds on the skills developed in the first one.</xhtml:p> </xhtml:div> </mlo:assessment> <mlo:objective>Have you ever thought of writing short stories or trying your hand at a novel? This 12-week online course provides a practical introduction to writing fiction - firing your creativity and imagination as well as equipping you with basic narrative strategies. During the course you will write two short pieces, for which your tutor will provide detailed individual feedback. You will also have the opportunity to work with your tutor, along with other students, in an online environment. You will read and learn from the works of writers as well as listening to their advice on beginning to write fiction.</mlo:objective> <mlo:prerequisite>There are no entry requirements but as this is an online course you should feel comfortable using a computer and accessing the internet. If you have any doubt about the suitability of the course, please contact our Student Registration & Enquiry Service.</mlo:prerequisite> <mlo:qualification> <dc:title>May lead to a variety of qualifications</dc:title> <dcterms:type>NLD Qual Code: M</dcterms:type> <awardedBy>The Open University</awardedBy> </mlo:qualification> <mlo:credit> <credit:scheme>The Open University</credit:scheme> <credit:level>1</credit:level> <credit:value>10</credit:value> </mlo:credit> <mlo:credit> <credit:scheme>SCQF</credit:scheme> <credit:level>7</credit:level> <credit:value>10</credit:value> </mlo:credit> <presentation> <dcterms:description>Year of entry: 2011</dcterms:description> <dc:identifier>A1742011E</dc:identifier> <mlo:start>2011-05-07</mlo:start> <end>2011-08-31</end> <mlo:duration interval="P4M">4 months (approx); NLD Duration Code: F</mlo:duration> <applyFrom>2010-09-05</applyFrom> <applyUntil>2011-04-08</applyUntil> <applyTo>http://www.open.ac.uk/study/</applyTo> <studyMode>NLD Study Mode Code: C</studyMode> <attendanceMode>Self-study</attendanceMode> <attendancePattern>Part time distance education course</attendancePattern> <mlo:languageOfInstruction>EN</mlo:languageOfInstruction> <languageOfAssessment>EN</languageOfAssessment> <mlo:cost>UK fee: GBP155.00; Ireland fee: GBP340.00; Continental Europe fee: GBP385.00</mlo:cost> </presentation> </course> <mlo:location> <mlo:street>PO Box 197</mlo:street> <mlo:town>Milton Keynes</mlo:town> <mlo:postcode>MK7 6BJ</mlo:postcode> <mlo:address>PO Box 197, Milton Keynes, MK7 6BJ</mlo:address> <mlo:phone>0845 3006090</mlo:phone> <mlo:fax>01908 653744</mlo:fax> <mlo:email>info@myinstitution.ac.uk</mlo:email> </mlo:location> </provider> </catalog> </code> Craig. https://profiles.google.com/matt.southall on "Data Definitions Example Problem?" http://www.xcri.org/forum/topic.php?id=197#post-817 Fri, 04 May 2012 12:57:39 +0000 https://profiles.google.com/matt.southall 817@http://www.xcri.org/forum/ Hi, Looking at the document: <a href="http://www.xcri.co.uk/KbLibrary/XCRI_CAP_Data_Definitions3.0.docx" rel="nofollow">www.xcri.co.uk/KbLibrary/XCRI_CAP_Data_Definitions3.0.docx</a> It shows that you can have a course identifier of the format: <dc:identifier xsi:type="http://www.ucas.ac.uk/:UCAScoursecode">V100</dc:identifier> Yet the validator at <a href="http://validator.xcri.co.uk/" rel="nofollow">http://validator.xcri.co.uk/</a> tells me that I cannot have a colon in the xsi:type there: "The attribute 'type' has an invalid value 'http://www.ucas.ac.uk/:UCAScoursecode' according to its schema type 'QName' - The ':' character, hexadecimal value 0x3A, cannot be included in a name. (1 failed instance)" Can anybody tell me which of these is correct? Or perhaps that I've misunderstood something here? Regards, Matt https://profiles.google.com/matt.southall on "OpenID Login Field Too Short" http://www.xcri.org/forum/topic.php?id=196#post-816 Fri, 04 May 2012 12:32:50 +0000 https://profiles.google.com/matt.southall 816@http://www.xcri.org/forum/ Hi, The OpenID login box up top on this forum is slightly too short to fit my OpenID URL: <a href="https://profiles.google.com/matt.southall" rel="nofollow">https://profiles.google.com/matt.southall</a> Luckily, it was exactly 1 character too short, and I managed to fudge it by using http instead of https - not ideal. Suggest raising the limit to something more like 255 characters. Regards, Matt Alan Paull on "Data Definitions and Vocabulary Framework documents" http://www.xcri.org/forum/topic.php?id=195#post-815 Thu, 03 May 2012 15:19:42 +0000 Alan Paull 815@http://www.xcri.org/forum/ CORRECTION TO LINK: <a href="http://www.xcri.co.uk/data-definitions-and-vocabulary-framework.html" rel="nofollow">http://www.xcri.co.uk/data-definitions-and-vocabulary-framework.html</a> Unfortunately a full stop at the end got included in error. Apologies. Alan https://profiles.google.com/106151444538427901807 on "StudyMode - sandwich?" http://www.xcri.org/forum/topic.php?id=194#post-814 Thu, 19 Apr 2012 07:35:09 +0000 https://profiles.google.com/106151444538427901807 814@http://www.xcri.org/forum/ HI Alan, apologies for such a late response. I understand your hesitation for the inclusion of a Sandwich StudyMode. If we can specify that there is an 'Optional Sandwich Mode' in the attendanceMode element then that will be fine and I believe we are allowed multiple values. Craig Hawker on "Data Definitions and Vocabulary Framework documents" http://www.xcri.org/forum/topic.php?id=195#post-813 Wed, 18 Apr 2012 19:19:13 +0000 Craig Hawker 813@http://www.xcri.org/forum/ The validator at <a href="http://validator.xcri.co.uk" rel="nofollow">http://validator.xcri.co.uk</a> has been updated with the additional guidance within the Data Definitions document, as well as having the ability added to validate against the various VDEX vocabularies provided the Course Data Programme namespace and XSD are referenced. If anyone encounters any issues with the validation of these new data definitions and vocabularies, please post then within the "Using the Validator" ( <a href="http://www.xcri.org/forum/forum.php?id=4" rel="nofollow">http://www.xcri.org/forum/forum.php?id=4</a> ) forum. Craig. Tavis on "Widgets for small-scale course information syndication" http://www.xcri.org/forum/topic.php?id=82#post-812 Wed, 18 Apr 2012 00:24:45 +0000 Tavis 812@http://www.xcri.org/forum/ I have posted some <a href="http://www.xcri.org/wiki/index.php/Sample_ASP.NET_and_XSLT_code_to_server_both_XHTML_web_pages_and_XCRI_CAP_1.2_XML_from_a_single_course_discovery_page">Sample ASP.NET and XSLT code to server both XHTML web pages and XCRI CAP 1.2 XML from a single course discovery page</a> on the wiki. If you request <a href="http://www.kelpiecollege.org.uk/courses/default2.aspx">this course discovery page</a> in a browser (with optional filtering parameters supplied through links), you will get a HTML web page, but if you request it with a "application/xcri-cap+xml" HTTP ACCEPT value, you will get an XCRI CAP 1.2 subset. This could be used to feed a remote widget. Craig Hawker on "Data Definitions and Vocabulary Framework documents" http://www.xcri.org/forum/topic.php?id=195#post-811 Fri, 06 Apr 2012 07:33:17 +0000 Craig Hawker 811@http://www.xcri.org/forum/ Just so everyone's aware: the validator is in the process of being updated with the additional rules contained within v3.0 of both the Data Definitions and Vocabulary Framework documents. Craig. Alan Paull on "Data Definitions and Vocabulary Framework documents" http://www.xcri.org/forum/topic.php?id=195#post-810 Thu, 05 Apr 2012 14:28:47 +0000 Alan Paull 810@http://www.xcri.org/forum/ Dear Colleagues Data Definitions and Vocabularies --------------------------------- The first version controlled release of the Data Definitions document (v3.0) and the Vocabulary Framework document (also v3.0) for the JISC-funded Course Data Programme are now available from the XCRI Knowledge Base at: <a href="http://www.xcri.co.uk/data-definitions-and-vocabulary-framework.html." rel="nofollow">http://www.xcri.co.uk/data-definitions-and-vocabulary-framework.html.</a> The link is a page of the Knowledge Base that explains the purpose of the two documents and provides links for downloading them. Both documents have been provided as Word files (.docx); there are hyperlinks in the document for internal navigation and to external resources. A simplified spreadsheet version of the Data Definitions document is currently in preparation and will be available next week. If you would prefer a different format, please contact me; it's likely that we will be able to accommodate other formats relatively easily and quickly. To give feedback on either of these documents please contact Alan Paull, <a href="mailto:alan@alanpaull.co.uk.">alan@alanpaull.co.uk.</a> Best wishes, and have a good holiday. Alan Paull <a href="mailto:alan@alanpaull.co.uk">alan@alanpaull.co.uk</a> Alan Paull on "StudyMode - sandwich?" http://www.xcri.org/forum/topic.php?id=194#post-809 Thu, 05 Apr 2012 09:56:46 +0000 Alan Paull 809@http://www.xcri.org/forum/ Unfortunately it's not quite that simple. studyMode vocabulary entries are currently mutually exclusive, so that searching can be carried out easily. Introducing SW in that way would mean that a search for Full time courses would exclude the SW ones, even though the study is actually Full time (studyMode is not repeatable). If we want to include SW in the studyMode vocabulary, we'd have to do something like: FT = Full time excluding Sandwich courses SW = Full time thick or thin Sandwich courses This would mean that implementers MUST use FT and SW to search for all Full time provision. I'm coming round to the view that SW is not a studyMode (which is to do with loading on the student), but rather a type of attendanceMode. Therefore attendanceMode might change to: CM Campus SW Campus-based with thick or thin sandwich DA Distance with attendance DS Distance without attendance NC Face-to-face non-campus MM Mixed mode ON Online (no attendance) WB Work-based I'd welcome views on this. Alan https://profiles.google.com/106151444538427901807/about?hl=en on "StudyMode - sandwich?" http://www.xcri.org/forum/topic.php?id=194#post-808 Tue, 03 Apr 2012 09:04:35 +0000 https://profiles.google.com/106151444538427901807/about?hl=en 808@http://www.xcri.org/forum/ Hi all As described in v.2 of the Data Definitions document, the studyMode element does not mention courses taught on a Sandwich Mode. At the University of Greenwich this is a possible offering. Therefore it will have a separate presentation with the studyMode as: <studyMode identifier="SW">Sandwich</studyMode> I assume this is okay? but are we the only ones doing this? and i will interested to know how other institutions offering a SW mode are handling this. https://profiles.google.com/106151444538427901807/about?hl=en on "Designing an 'XCRI' Data Model" http://www.xcri.org/forum/topic.php?id=191#post-807 Mon, 02 Apr 2012 12:10:21 +0000 https://profiles.google.com/106151444538427901807/about?hl=en 807@http://www.xcri.org/forum/ Tavis, thank you for making your content available. It is certainly very interesting to see how institutions have modeled there content and the rational behind there decisions. I am not sure how many BANNER universities are participating in this XCRI project, but before we make a decision in regards to implementing a brand new structure, we will first look into using existing tables and/ or see if BANNER has anything which resembles what we are looking for and can be modded. For now, we have modeled a "draft" brand new data model which compliments and resembles the XCRI model closely. Looking briefly at your model the core structure is very similar, but we have yet to make ours as extensible and future proof as yours. As soon as we have something slightly more concrete I will make what we have available and happily share any ideas or thoughts on your model if I spot anything. I think that our main area of concern in regards to a new model is the initial loading of data, maintenance, synchronization and building a front-end. Tavis on "Designing an 'XCRI' Data Model" http://www.xcri.org/forum/topic.php?id=191#post-806 Sun, 01 Apr 2012 21:23:23 +0000 Tavis 806@http://www.xcri.org/forum/ I've posted my development database design and code up on the wiki under <a href="http://www.xcri.org/wiki/index.php/XCRI_Wiki#Code_and_Examples_for_XCRI-CAP_1.2">Code and Examples for XCRI-CAP 1.2</a>. The design tries to address issues behind some of the comments made here, about extensibility/future-proofing and generalization, and some parts of our solutions are likely to be similar. As well as the data diagram, physical schema code and sample data, there's a stored procedure (and a function) which returns a valid XCRI CAP 1.2 XML output. This is a work in progress and doesn't yet meet all Craig's validator rules, but it is fairly comprehensive. I'm not directly involved in the Course Data Programme, so that's likely where some omissions will lie. It's all under the Wiki's Attribution 3.0 licence, so take anything you feel like, and please let me know if you spot a problem with anything (it hasn't been thoroughly tested). I needed this database anyway before moving on to (personal) web development work. I think that most institutions will have existing, usable database tables, and might be looking to supplement them with additional data sources, so perhaps a range of hybrid solutions will emerge, which will be interesting for comparison's sake. Alan Paull on "Guidance sought on data mapping" http://www.xcri.org/forum/topic.php?id=188#post-805 Fri, 30 Mar 2012 10:25:28 +0000 Alan Paull 805@http://www.xcri.org/forum/ Yes, that's a good summary. I would recommend having a look at Big XCRI (XCRI r1.0)for more general courses information modelling. It's intended to cover "everything about courses" and it maps to XCRI-CAP. The schema is at: <a href="http://www.xcri.co.uk/schemas/xcri_r1.0.xsd." rel="nofollow">http://www.xcri.co.uk/schemas/xcri_r1.0.xsd.</a> And there's more information at: <a href="http://www.elframework.org/projects/xcri." rel="nofollow">http://www.elframework.org/projects/xcri.</a> It was used in the CUMULUS project: <a href="http://www.xcri.co.uk/h2-mm-resources-a-z/259-cumulus-project.html." rel="nofollow">http://www.xcri.co.uk/h2-mm-resources-a-z/259-cumulus-project.html.</a> I have some hopefully useful slides on r1.0 if anyone is interested. Alan http://qamar-xaman.myopenid.com/ on "Guidance sought on data mapping" http://www.xcri.org/forum/topic.php?id=188#post-804 Fri, 30 Mar 2012 07:44:40 +0000 http://qamar-xaman.myopenid.com/ 804@http://www.xcri.org/forum/ In my limited experience of only a few instititions, teh problem is not with XCRI-CAP and what it is designed to achieve; its rather the intent of teh institutions and what they want to see it solve. A lot of institutions do not yet have a central system/database/model to hold course data. By introducing XCRI CAP in their institutions, they want to 'get their house in order', as it gives them a chance to look at their data strategy and design something from grounds up. The obvious impression is why not model something based on the same schema as Xcri, however Xcri schema is not intended for course modelling, I guess. So these institutions need to model their 'course input/management systems' independant of Xcri schema. So in other words, Xcri format should only be thought of as syndicating a subset of course information to specific partners. The information held within the document and its structure is driven by those partners and not by the institutions? At Harper-Adams, my original assumption was to create a data model that is based on Xcri schema, but over time I am more inclined to create a model independant of Xcri, with only the knowledge that it is able to generate a valid Xcri feed. However, my main fear is that in order to share course information with other universities, I may need to come up with 'yet another' schema to satisfy the level and detail of information required by them. Is this a fair assessment? Tavis on "Guidance sought on data mapping" http://www.xcri.org/forum/topic.php?id=188#post-803 Thu, 29 Mar 2012 17:36:34 +0000 Tavis 803@http://www.xcri.org/forum/ If the requirement is just to group courses into chapters of a prospectus or a navigational hierarchy of a website, you could use the <a href="http://www.xcri.org/wiki/index.php/XCRI_CAP_1.2#the_.3Csubject.3E_element">subject element</a>. Our college uses a subject-based classification we call "cluster > segment > segment section", which we use for our <a href="http://www.adamsmith.ac.uk/courses/Pages/default.aspx">course webpages</a> and prospectus, and we could have used a subject element for that. In fact I think we used an extension to older XCRI Curriculum and linked the identifiers to an external XML file which held more data about the class, such as colour scheme and so on, where the hierarchical relationship of the classes was represented. Alan Paull on "Guidance sought on data mapping" http://www.xcri.org/forum/topic.php?id=188#post-802 Thu, 29 Mar 2012 15:10:19 +0000 Alan Paull 802@http://www.xcri.org/forum/ I echo what Scott said above, in that this was our initial thinking with the XCRI Course Advertising Profile, as opposed to Big XCRI, which covers pretty much anything. There are parts of our specification landscape that are fixed at the moment (within XCRI-CAP 1.2 / BS 8581 to be). Nesting providers is not possible. I would also say that our early researches and pilot implementations didn't come up with a requirement for this for the purposes of course advertising. This type of organisational description *could* be included in provider>>description: "This degree programme is offered by the School of Computing, which is part of our world-renowned Faculty of Technology" or some such. It is permitted within XCRI-CAP 1.2 to have multiple providers off a single catalog. However, in the interests of keeping it simple, the current Data Defs draft permits only one provider in a catalog, so that on aggregation the representation of the HEI is clear. With multiple providers representing organisational sub-units, there is a danger that the HEI name itself will not be included, or that a consuming service won't know how to handle the data properly. One possible solution suggests itself, though I would hesitate to recommend it at this stage, as we're very close to the first release of the data defs - so this may be an idea for the future. It might be possible to use multiple provider elements, requiring that the first element is the HEI. This first provider element could theoretically contain no courses (permitted in XCRI-CAP, but not really expected). Subsequent provider elements could show the major organisational sub-components at the first level of the hierarchy (for example Faculty) with each one containing its courses. Then consuming systems could be required to treat the first provider element as the HEI, with subsequent ones processed as organisational components. I think that this might be overkill for XCRI-CAP usage, and I personally prefer putting the marketing text in the provider >>description as a solution. We're apt to fudge the issue to general dissatisfaction, if we try to bend XCRI-CAP. An alternative solution is to add in a <description xsi:type="childOrganisation">, which could be applied to <provider>, <course> or even <presentation> if needed. Use of this description type could be constrained by a definition in the Course Data Programme schema (for example to have space only for the name of the sub-unit), so it would still meet the XCRI-CAP requirements, but fulfil an additional requirement for the name of a school, faculty or department. Alan Scott Wilson on "Guidance sought on data mapping" http://www.xcri.org/forum/topic.php?id=188#post-801 Thu, 29 Mar 2012 09:11:11 +0000 Scott Wilson 801@http://www.xcri.org/forum/ I think when we did our original research we felt that exposing the internal structure of the University wasn't a major concern when advertising courses to prospective students. In most cases the relevant aspects could be covered in descriptive marketing text rather than modelling the structure. Organisation modelling is more important when reusing XCRI for curriculum design and modelling purposes https://www.google.com/accounts/o8/id?id=AItOawl3uO8Ytv2_hcifnqqDD352mb5qceH7V1U on "Guidance sought on data mapping" http://www.xcri.org/forum/topic.php?id=188#post-800 Thu, 29 Mar 2012 08:21:39 +0000 https://www.google.com/accounts/o8/id?id=AItOawl3uO8Ytv2_hcifnqqDD352mb5qceH7V1U 800@http://www.xcri.org/forum/ But <code><catalog></code> is only allowed to contain one provider (p15 Data Defs 2.0). What I'd like to see is: <pre><code><catalog> <provider> <provider> <course></course> </provider> <provider> <course></course> </provider> ..... </provider> </catalog></code></pre> The outer provider is University of Oxford, the inner providers are departments. The could be further extended to help match the structure of our Universities so that "teaching units" could be represented as <code>University > Campus > Division > Department > Sub-department</code> Other Universities have different structures such as <code>University > Faculty > Department</code> After doing a bit of research last week it transpires that all the UK institutions who run Sakai have a different hierarchical structure in place with a different number of levels and different names. I suppose what I am proposing is allowing arbitrarily nested providers whilst at the same time changing the content of <<code>provider></code> so it must contain either one or more <code><course></code> elements or one or more <code><provider></code> elements. Adam http://qamar-xaman.myopenid.com/ on "Guidance sought on data mapping" http://www.xcri.org/forum/topic.php?id=188#post-799 Wed, 28 Mar 2012 09:16:10 +0000 http://qamar-xaman.myopenid.com/ 799@http://www.xcri.org/forum/ Taking a step back, here is a slightly different take on this... Catalog is at the highest level. If we assume catalog as the container to put all courses for an institution, then Catalog becomes this virtual entity (the institution). It has all of the common elements (title, Resource Identifier, etc.) to describe it. Provider is then defined as a venue/campus entity. If an institution like Oxford University has multiple campuses, then each becomes a provider, and course entity can either sit under the provider or under catalog alongside the provider, and in the latter cases, venue for a presentation will determine where it is taught etc. <catalog> <provider> <course></course> </provider> <course></course> </catalog> I know this is not the intent of the catalog element, but can work as a possible solution that requires nominal change to the structure... any thoughts? https://www.google.com/accounts/o8/id?id=AItOawl3uO8Ytv2_hcifnqqDD352mb5qceH7V1U on "Guidance sought on data mapping" http://www.xcri.org/forum/topic.php?id=188#post-798 Wed, 28 Mar 2012 08:37:01 +0000 https://www.google.com/accounts/o8/id?id=AItOawl3uO8Ytv2_hcifnqqDD352mb5qceH7V1U 798@http://www.xcri.org/forum/ I was going to say that but Alex beat me to it! the Bodleian Libraries give many of their courses in the OUCS building because they don't have suitable lecture rooms. https://www.google.com/accounts/o8/id?id=AItOawmcxfq9NUT2pS0gqg9hO4Fy2yZrLuv9vRY on "Guidance sought on data mapping" http://www.xcri.org/forum/topic.php?id=188#post-797 Wed, 28 Mar 2012 08:08:13 +0000 https://www.google.com/accounts/o8/id?id=AItOawmcxfq9NUT2pS0gqg9hO4Fy2yZrLuv9vRY 797@http://www.xcri.org/forum/ The draft version of the spec available from <a href="http://www.xcri.co.uk/data-definitions-and-vocabulary-framework.html" rel="nofollow">http://www.xcri.co.uk/data-definitions-and-vocabulary-framework.html</a> says: """The venue element SHOULD be used to describe the physical venue of the presentation. This MAY be a sub-organisation of the provider of the presentation (such as a department or school), or it MAY be a third party or external organisation.""" This conflates the organisational unit and the locations it occupies (an organisation is a social institution; it has no physical presence. Yes, they both have addresses, but address of an organisation is for correspondence, whereas that of a building is to physically locate it). I've taken venue/provider to mean "physical venue", and in our current output we ignore any organisational overtones of the element. We have a many-to-many relationship between organisational parts and the spaces they occupy (and indeed, we have training provided by organisational parts in spaces they don't otherwise occupy). All this means that the address of the provider could be misleading as a location for the venue, and I'd far rather they be split out like this: <pre><code><presentation> <venue> <dc:title>3 James Street</dc:title> <mlo:address>…</mlo:address> … </venue> <provider> <dc:title>Department of Computer Science</dc:title> <mlo:address>…</mlo:address> </provider> </presentation></code></pre> Multiple venues would be catered for by repeating the venue element, and joint schools (Maths and Computer Science, for example) could have multiple providers. Craig Hawker on "Designing an 'XCRI' Data Model" http://www.xcri.org/forum/topic.php?id=191#post-796 Tue, 27 Mar 2012 17:19:00 +0000 Craig Hawker 796@http://www.xcri.org/forum/ Hi Qamar, The source code is downloadable from the project repository at <a href="http://code.google.com/p/xcricap-net/." rel="nofollow">http://code.google.com/p/xcricap-net/.</a> You will need a Mercurial client to pull/push code; I can give you some guidance on that if you haven't used it before. Get my email address from your Course Data Programme representative and we'll sort it off-list. Craig. Craig Hawker on "Using the 1.2 validator" http://www.xcri.org/forum/topic.php?id=184#post-795 Tue, 27 Mar 2012 17:16:19 +0000 Craig Hawker 795@http://www.xcri.org/forum/ Hi Qamar, Get my email address from your Course Data Programme representative (let me know if you have any problems finding out who's helping Harper-Adams and I can find out from this end). I'm more than happy to commit some time but it will be evenings and weekends only I'm afraid. Craig. http://qamar-xaman.myopenid.com/ on "Using the 1.2 validator" http://www.xcri.org/forum/topic.php?id=184#post-794 Tue, 27 Mar 2012 16:56:31 +0000 http://qamar-xaman.myopenid.com/ 794@http://www.xcri.org/forum/ Thanks Craig, I don't definately want to take up much more of your time, but felt that I could contribute and help the community with further changes that may be required. (I am already happy with the detailed comments you provided). But if you do have time at some point, please remember that you have at least one soul that is willing to contribute! Thanx Qamar http://qamar-xaman.myopenid.com/ on "Designing an 'XCRI' Data Model" http://www.xcri.org/forum/topic.php?id=191#post-793 Tue, 27 Mar 2012 16:50:00 +0000 http://qamar-xaman.myopenid.com/ 793@http://www.xcri.org/forum/ Craig - Thanks ever so much for the update. I have access to the DLL, but not the Source Code. As we are at the stage of writing something (purely for and by the Educational institution), it would be good for us to start with that and develop it further with a view to making it available to everyone else - essentially contributing to your project under your existing setup/license etc. I am not sure how to go about it and whether we need to wait for 1.2 standard to stabilize over the next couple of weeks. Regardless, we are happy to start contributing ASAP. Mehmet - I wanted to explain this to others who may be facing similar situations, hence my discussion on UI and how data would be entered. You can definitely make it complex at the database level, but that later on can create its challenges in designing the data input screens…. Hence my discussions on trying to tackle this at a broader level. Here is a quick breakdown of one of ways we chose to model it in SQL Server. • There are essentially four main entities (Catalog, Provider, Course, and Presentation). Each is a one-to-many to the previous one. • In my model, I’ve chosen to model Descriptive common text elements as separate tables with one-to-many with Course and Presentation. This gives the extra 14 tables; CourseObjective, PresentationObjective, CourseAssessment and so on. Taking objective as an example, if you only have one objective element per course, then there is no need to have a separate table for this. However Xcri format allows you to have multiple objectives per course. At the database level, I’ve chosen to implement each objective as a separate row in a CourseObjective table, with a foreign-key relation back to the course. If stored at this granular level, it gives the most flexibility in how you want to combine this and present to the user, but may be an overkill for most people. • Rather than having a separate table per descriptive element , a single table that holds a value and a type may also suffice. • Qualifications and credits are many-to-many relations with course and will require a link table in between for each of these. I also had a number of lookup tables for AttendanceMode, AttendancePattern and StudyMode. I also did another prototype where I used inheritence of tables by creating shared Common Elements and Common Descritive Element tables, but again an overkill and unnecessary complexity. I personally believe four core tables, plus the Qualifucations/Credit tables (only if needed) is all that ought to be required. Qamar Craig Hawker on "Using the 1.2 validator" http://www.xcri.org/forum/topic.php?id=184#post-792 Tue, 27 Mar 2012 16:24:09 +0000 Craig Hawker 792@http://www.xcri.org/forum/ Hi Qamar, The purpose of the library was to be just that: a starting point for institutions to pick up. It was not expected to be a complete solution but something that takes the grunt heavy-lifting away and means they don't have to concern themselves with ensuring correct namespaces and the like. From the work I've done on validation it's simply things like namespace prefixes (and declarations) and element casing which are the most frequently broken. The generator library has taken a back seat. Initially this was due to the fact that I was struggling to validate XCRI-CAP 1.2 data due to some issues with the schemas but continued primarily because institutions didn't seem to be interested in using it. Thus far, aside from this post, only Jade (above) has contacted me to say they're using it and to submit some patches. So it hasn't been actively (or otherwise!) developed in about 12 months, from my memory of the last check-in. That said, aside from any issues or extension work that comes up as it gets used further, I have just finished up with the validator. As a result I am more than willing to put some time into the generator libraries again. If anyone would like to contribute further then get my email address from your Course Data Programme representative (let me know if you have a problem finding out who this is) and I'll get you set up as a contributor. I know that you raised a question regarding the licensing in another thread but this caveat is solely there to stop commercial entities picking it up and charging institutions for it. It's free for staff at universities/colleges/etcetera to use as they wish. If that's something that you want to change then let me know and we can discuss it off-list. I thought that there were concrete implementations of all elements but I may well be wrong. It's been a long time since I looked and I am fully aware that the test library (which may not be part of the zip, I can't remember) is nowhere near good enough. I will check the specific elements you mentioned when I get a chance but it sounds like what you did would be correct. The output from the generator aims to be valid XCRI 1.2, to the standard. However it mainly aims to ensure the structure and namespaces are correct - I purposefully avoided adding any element of validation to the library. Therefore it should output the correct elements in the correct order but it will not complain if you have ommited mandatory elements. The last check-in was way before the 1.2 validator was produced so it has not been tested with this. Also bear in mind that the 1.2 validator additionally validates according to the Data Definitions Document (currently v1.6, with 2.0 being drafted). Basically: it was aimed to be a starting point but things have progressed a lot since the last check-in. I am willing to put some more time into it (albeit in my own time, so bear that in mind), even if that is just to manage the repository. If you'd like to contribute then I'll welcome you with open arms. :-) Craig. Craig Hawker on "Designing an 'XCRI' Data Model" http://www.xcri.org/forum/topic.php?id=191#post-791 Tue, 27 Mar 2012 15:57:07 +0000 Craig Hawker 791@http://www.xcri.org/forum/ Everyone, I don't want to hijack the thread but I wanted to reply to Qamar's comments about the generator library. Please note that the project is open-source and if anyone would like to submit patches for the elements they find incomplete to help the rest of the community then please let me know and I'll add you to the list of contributors. To this date only a single person has contacted me to say they've used the library and that one person submitted a patch for a few bits. Even submitting issues via the project would be beneficial. Specifically: 1. The licence allows for free use by any educational institution with an attribution. The only restriction is on commercial entities lifting it and charging institutions for its use. 2. Qamar's right about it being dated. The modifications for 1.2 were done back when the 1.2 standard was being developed. Some time has passed and some of the guidance has changed. Equally, for the work that's ongoing at the moment, the data definitions document that guide institutions on what to produce is still going through versioning so it's still a bit up in the air about what gets implemented and when. 3. If there's something I can do to change the project to help any perceived lack of control over future changes then just let me know. However there's nothing, currently, to stop an institution cloning the project now and customising it to exactly fit their needs. Alternatively any issues could be reported (and/or patched) so everyone could use them, and the project just extended within an institution-specific DLL. The point of this project was to try and give institutions an alternative to having to write everything themselves. Now that the validator is available I may well end up with some spare time to look at any reported issues. Whether that gets done in time to help the institutions currently implementing feeds, I don't know... Craig. https://profiles.google.com/106151444538427901807/about?hl=en on "Designing an 'XCRI' Data Model" http://www.xcri.org/forum/topic.php?id=191#post-790 Tue, 27 Mar 2012 13:09:40 +0000 https://profiles.google.com/106151444538427901807/about?hl=en 790@http://www.xcri.org/forum/ Qamar, thank you for sharing your thoughts and decisions. As I mentioned before we are in a similar situation where course information is dispersed in various other systems, with the majority held in our Marking database. Our decision is to create a structure to store all course information within our main Student Record Systems database, therefore our first hurdle is considering how we are going to structure our course information within Oracle and allow for complete flexibility in terms of holding similar course information “repeating attributes” at different levels. I appreciate that considering a "front-end" for data population and maintenance is a vital consideration which cannot be overlooked in the early stages but we have not thought about it in the depth with Harper Adams have. We have briefly discussed the possibilities and options we have. But your ideas will certainly come helpful later down the line. For the Initial load we will build a mechanism to export the data from marketing and import into our new structure. Regarding data-models, we still need to make a decision as to at what level data elements can change i.e. can there be a different block of text for the ‘pre-requisite’ data element for different modes, years and start month’s of a course? I think I will try to design a model that is flexible in terms of adding different content for a course by year, mode and possibly start month. Trying keep it future proof. In order to produce a “presentation” for every offering of a course. An initial table stating the Course structure, term, modes and start months could be used. So for the below course, we would have four different presentations layers for COURSE1. For example: <code> 1_UP_ID COURSE YEAR MODE START 1 COURSE1 2012 PT JAN 2 COURSE1 2012 PT SEP 3 COURSE1 2012 FT JAN 4 COURSE1 2012 FT SEP </code> Another table could use the 1_UP_ID to identify data elements at that level. So for the part-time, January starter, COURSE 1 presentation level we can have custom text for objective, pre-requisite, and application procedure text. <code> FK_1_UP_ID ELEMENT CONTENT Start date End date 1 OBJECTIVES Some text 2012 99999 1 PREREQUISITES Some text 2012 99999 1 APPLICATION_PROC Some text 2012 99999 </code> This allows us to store different text at presentation layer. For the COURSE level we would just have a COURSE table which has all the seven data elements and text. <code> COURSE ELEMENT CONTENT Effec_year COURSE1 ABSTRACT Some text 2012 COURSE1 PREREQUISITES Some text 2012 COURSE1 APPLICATION_PROC Some text 2012 e.t.c e.t.c e.t.c e.t.c </code> I certainly have not explained this as well as I could have, but this a draft so far. Please allow me to elaborate if I have made myself unclear. I would certainly be interested to see your data model and SQL Qamar. My email is <a href="mailto:m.batmaz@gre.ac.uk">m.batmaz@gre.ac.uk</a> http://qamar-xaman.myopenid.com/ on "Related Courses" http://www.xcri.org/forum/topic.php?id=193#post-789 Tue, 27 Mar 2012 12:39:10 +0000 http://qamar-xaman.myopenid.com/ 789@http://www.xcri.org/forum/ Thanks Scott, There was some interest shown in the webinar two weeks ago, but I think the interest will come from aggregators and UI developers who would want to provide a richer experience. Consuming data 'linearly' may not promote much interest (e.g. Click on Provider to see all cources, click on a course to see all presentations), but a more interactive presentation of courses will require these relations, either inherent in the schema, or built dynamically. Well, lets wait and see what others have to say. Q Scott Wilson on "Guidance sought on data mapping" http://www.xcri.org/forum/topic.php?id=188#post-788 Tue, 27 Mar 2012 12:35:36 +0000 Scott Wilson 788@http://www.xcri.org/forum/ The easiest approach is to use the <provider> element as a child of presentation -> venue as its designed for this purpose. From the section on <venue>: <ul> <li>Provider: Producers SHOULD use the provider element to describe the organisation which acts as the provider of the venue. This MAY be a sub-organisation of the provider of the presentation (such as a department or school), or it MAY be a third party or external organisation.</li> <li>Provider Properties: When a provider element is used in a venue element, Producers SHOULD include only basic information about an organisation, such as identifier, title, description, url, image and location.</li> </ul> (That last guideline is really to discourage "nesting" from a venue, rather than discourage providing useful information about the venue) Scott Wilson on "Related Courses" http://www.xcri.org/forum/topic.php?id=193#post-787 Tue, 27 Mar 2012 12:26:49 +0000 Scott Wilson 787@http://www.xcri.org/forum/ Hi Qamar, Relationships between courses (or rather, between modules and programme elements) was also part of the curriculum modelling aspect of the original "Big XCRI"; this hasn't been developed in recent years primarily as there hasn't been a critical mass of institutions wanting to try it out. However if there is renewed interest in modelling course relationships then we'd be happy to provide some wiki space on xcri.org for developing it (and of course you can use this forum). S (PS We turned off allowing new memberships on the wiki recently because of spam, but will be back up once we've installed a new spam filter) https://www.google.com/accounts/o8/id?id=AItOawl3uO8Ytv2_hcifnqqDD352mb5qceH7V1U on "Guidance sought on data mapping" http://www.xcri.org/forum/topic.php?id=188#post-786 Tue, 27 Mar 2012 12:24:54 +0000 https://www.google.com/accounts/o8/id?id=AItOawl3uO8Ytv2_hcifnqqDD352mb5qceH7V1U 786@http://www.xcri.org/forum/ It would be most useful to be allowed to have (1) a <provider> within the <course> element and possibly (2) one within the <presentation> which could be inherited from <course>. The thinking behind (1) is that for graduate training, iot is vitally important to know which department is offering the course. We have scores of courses called "Research Methods" as a student you want to attend the course from your home department as it will be the most relevant. The thinking behind (2) is that the actual presentation may be done by a different department 9although I'm not that convinced by that argument!) http://qamar-xaman.myopenid.com/ on "Related Courses" http://www.xcri.org/forum/topic.php?id=193#post-785 Tue, 27 Mar 2012 12:11:29 +0000 http://qamar-xaman.myopenid.com/ 785@http://www.xcri.org/forum/ We sort of talked about this ona previous webinar (14 March). Essentially I wanted the ability to link cources to other courses - and there could be a host of reasons why. There were other people interested in this as well. Websites that will aggregate courses will also find it useful. Here are some thoughts on how that and other types of links may be implemented. I think we need something like the following 'relatedElements' complex element that can be added to any of the main types; <relatedElements relatedTo="Course|Provider|Presentation" type="isPartOf|location|prerequisite|"> <ResourceIdentifier>http://...</ResourceIdentifier> <ResourceIdentifier></ResourceIdentifier> <ResourceIdentifier></ResourceIdentifier> </relatedElement> Specific examples... - The following example shows a list of venues where a course is available - note that presentation have a venue element that can link back to a provider, but this is restricted to only the current location of that course offering (not all possible locations). However, here I want to highlight all locations where that course is being taught (or has been or will be taught at some point). This is a contrived example as there are alternative (correct) ways to map this relation - the most common way is through multiple presentations. But I can still see a need for it. <relatedElements relatedTo="Provider" type="location"> <ResourceIdentifier>http://harper-adams/alternative/campus/234567</ResourceIdentifier> </relatedElement> - Course Pre-requisite cources <relatedElements relatedTo="course" type="requirement"> <ResourceIdentifier>http://harper-adams/courses/level1/PHY01</ResourceIdentifier> <ResourceIdentifier>http://harper-adams/courses/level1/BIO01</ResourceIdentifier> <ResourceIdentifier>http://harper-adams/courses/level1/CHE01</ResourceIdentifier> </relatedElement> Does anyone else have any need for any of 'relation' building within Xcri format? Qamar Xaman Harper Adams University College