Haven't Forgot DIVAS

No haven't forgotten about the DIVAS project. Just not much happening from my side - things have moved focus over to Sam and Shaun's side of things (technical). When we have something to play with - I will post a new.

XML Beans

Not much to post - but thought it pertinent to post that Shaun is using XML Beans to help the process of developing the web services Sam and Shaun are developing for uploading documents to HIVE and recording the associated data in XCRI-CAP.

"XMLBeans uses XML Schema to compile Java interfaces and classes that can then be used to access and modify XML instance data." Wikipedia

I am currently working developing a cut-down schematic of the XCRI-CAP diagram to outline what elements we are using from the schema. In addition, the details of drop-down lists (where applicable).

New Year! XCRI & HIVE Breakthrough

Happy 2009!!

So far so good - had an email from Sam to say that he has made a major breakthrough with developing an interactive schema for entering metadata against documents being uploaded to HIVE.

Sam says:

I've got Hive working with a revised version of XCRI CAP metadata. In this version, I've replaced all the XML attributes (Hive can’t cope with attributes) with XML elements. I've also added faculty as an extension of the course metadata and allowed more than one faculty to be associated with it. Attached is the schema definition if you want the details.

I've created and SDI file and wired it in so the metadata can be used in Hive. It's in and working but it's got a few usability issues. We could do with meeting up at some point to figure out how to refine the interface so it works best. If you want to have a look at the interface, just try adding an item of type "Validation document". The XCRI element-only metadata is associated with this type.

This is a modified version of the XCRI-CAP. I will look into it and test it's output. Shaun is currently working on an upload interface that assists users in uploading and 'marking up' documents - making the process a lot easier (hopefully).

Well done everyone!

DIVAS Meeting Update

Chris has provided a list of the following for drop-down lists:

• Professional and Statuatory Regulatory Bodies = 13 types
• Awards = 58 types

DIVAS Meeting Update

Present: Ben, Sam, Shaun, Chris & Sue.

We discussed what we will be recording in the metadata file when uploading validation documents to HIVE. In addition, we discussed the 'submission form' and some the process involved in uploading and searching for documents in HIVE.

This is what we decided:

Subject (XCRI = Subject)

• Discussed use of JACs
• Decided not to use this field/element at the moment

Level (XCRI = Level)

• Chris to provide a definitive list of levels for Drop-down / Look-up List
• Multiple instances recorded

Validation Date (XCRI = @generated)

• Date to be recorded = date of validation panel

Face-2-Face (XCRI = AttendenceMode)

• Drop-down / Look-up List:


1. Face-2-Face
2. Distance
3. Blended

Site of Delivery (XCRI = Venue)

• Multiple instances recorded

Form of Assessment (XCRI = AssessmentStrategy)

• Discuss it’s relevance for validation
• Could be useful for recording data for disability
• Decided not to use this field/element at the moment

Mode of Delivery (XCRI = StudyMode)

• Drop-down / Look-up List:

1. Full Time
2. Part Time
3. Fast Track 2 Year

• Multiple instances recorded

Faculty/School (XCRI = TBC)

• Need to investigate this further – where to record in XCRI
• Drop-down / Look-up List:

1. Faculty of Arts, Media and Design (AMD)
2. Business School
3. Faculty of Computing, Engineering and Technology (FCET)
4. Law School
5. Faculty of Health
6. Faculty of Science
7. Academic Development Institute (ADI)

Award Title (XCRI = Title)

• Record a String title unique to the document being uploaded
• Multiple instances recorded

Professional Body / Association (XCRI = AccreditedBy)

• Chris to provide a definitive list of levels for Drop-down / Look-up List

Recording Document Types

Possibly recorded in XCRI Description – unique to the individual documents being uploaded

Pre-Validation Document Types
1. Validation Support Descriptor
2. Rational
3. Programme Specification
4. Module Descriptor
5. Handbooks

a. Award / Student
b. Module
c. Placement
d. Mentor
e. *Sue to speak to R. Bennifer about other types for Work Based Learning

6. Resources

Post-Validation Document Tyes
7. Validation Report
8. Validation Report Response

The Validation Collection

• The documents are recorded as a collection in a ‘Catalogue’ folder in HIVE
• Currently the title of the collection / ‘Catalogue’ is largely heuristic (made up) by QIS – as evidenced in the shared QIS drive

Big Push Meeting

Have arranged a meeting between Sam & Shaun (technical / coding), myself and Chris (QI). We aim to thrash out the finer points of uploading documents to HIVE.

The main areas of interest are:

• Determining which parts of XCRI-CAP we are going to use / focus on
  
• How the validation documents are 'arranged' within HIVE - i.e use of collections & metadata & type of data to be recorded (doc type etc..)
  
• Discuss the technology required to upload documents with the metadata
• Exploring the search facility
  

Essentially, we are aiming to get documents into HIVE, be able to search for them (in context of the validation collection) and then presen them to the user.

DIVAS Meeting

Had a DIVAS meeting today - all productive. I was asked to supply a short paragraph / statement about my side of the project - so here it is (I will try and strip out as much of the teccy stuff as I can):

We had some problems matching the required document attributes with the element/fields of specific metadata schemas - such as Dublin Core and LOM. In essence, the DIVAS project needs to record some specific details - against which users could search for these documents. for example:

• Subject
• Level
• Form of Assessment
• Prof Associated Body etc....
  

Rather than finding a quick fix / imperfect workaround with the schemas that HIVE already uses - it was agreed that XCRI-CAP is a schema that matches very closely to what we were trying to record (see image).

Currently we are working on the HIVE repository being able to mark up documents stored in it against the XCRI-CAP metadata schema.

Essentially, HIVE requires an SDI file (XML file) to be written that tells HIVE how to generate a GUI (Graphical User Interface) - this user interface is a graphical representation of the metadata schema tree - providing the user with the field elements that make up the metadata tree and (associated with each element) the type of data you can record (see image - Dublin Core Example).

Sam has transformed the XCRI-CAP schema document (XSD file) into an SDI file by creating an XML instance of the XSD using XMLSpy. This has created a valid and fairly usable GUI in HIVE but the ouput XML file has a few errors - therefore some work needs to be done on correcting this.

Work continues and as long as we can generate a vliad output (XML output) that is valid for XCRI-CAP then it looks good to start uploading documents to HIVE.