OUseful.Info, the blog…

Trying to find useful things to do with emerging technologies in open education

Posts Tagged ‘eSTEeM

My ILI2012 Presentation – Derived Products from OpenLearn/OU XML Documents

FWIW, a copy of the slides I used in my ILI2012 presentation earlier this week – Making the most of structured content:data products from OpenLearn XML:

I guess this counts as a dissemination activity for my related eSTEeM project on course related custom search engines, since the work(?!) sort of evolved out of that idea…

The thesis is this:

  1. Course Units on OpenLearn are available as XML docs – a URL pointing to the XML version of a unit can be derived from the Moodle URL for the HTML version of the course; (the same is true of “closed” OU course materials). The OU machine uses the XML docs as a feedstock for a publication process that generates HTML views, ebook views, etc, etc of a course.
  2. We can treat XML docs as if they were database records; sets of structured XML elements can be viewed as if they define database tables; the values taken by the structured elements are like database table entries. Which is to say, we can treat each XML docs as a mini-database, or we we can trivially extract the data and pop it into a “proper”/”real” database.
  3. given a list of courses we can grab all the corresponding XML docs and build a big database of their contents; that is, a single database that contains records pulled from course XML docs.
  4. the sorts of things that we can pull out of a course include: links, images, glossary items, learning objectives, section and subsection headings;
  5. if we mine the (sub)section structure of a course from the XML, we can easily provide an interactive treemap version of the sections and subsections in a course; generating a Freemind mindmap document type, we can automatically generate course-section mindmap files that students can view – and annotate – in Freemind. We can also generate bespoke mindmaps, for example based on sections across OpenLearn courses that contain a particular search term.
  6. By disaggregating individual course units into “typed” elements or faceted components, and then reaggreating items of a similar class or type across all course units, we can provide faceted search across, as well as university wide “meta” view over, different classes of content. For example:
    • by aggregating learning objectives from across OpenLearn units, we can trivially create a search tool that provides a faceted search over just the learning objectives associated with each unit; the search returns learning outcomes associated with a search term and links to course units associated with those learning objectives; this might help in identifying reusable course elements based around reuse or extension of learning outcomes;
    • by aggregating glossary items from across OpenLearn units, we can trivially create a meta glossary for the whole of OpenLearn (or similarly across all OU courses). That is, we could produce a monolithic OpenLearn, or even OU wide, glossary; or maybe it’s useful to have redefine the same glossary terms using different definitions, rather than reuse the same definition(s) consistently across different courses? As with learning objectives, we can also create a search tool that provides a faceted search over just the glossary items associated with each unit; the search returns glossary items associated with a search term and links to course units associated with those glossary items;
    • by aggregating images from across OpenLearn units, we can trivially create a search tool that provides a faceted search over just the descriptions/captions of images associated with each unit; the search returns the images whose description/captions are associated with the search term and links to course units associated with those images. This disaggregation provides a direct way of search for images that have been published through OpenLearn. Rights information may also be available, allowing users to search for images that have been rights cleared, as well as openly licensed images.
  7. the original route in was the extraction of links from course units that could be used to seed custom search engines that search over resources referenced from a course. This could in principle also include books using Google book search.

I also briefly described an approach for appropriating Google custom search engine promotions as the basis for a search engine mediated course, something I think could be used in a sMoocH (search mediated MOOC hack). But then MOOCs as popularised have f**k all to do with innovation, don’t they, other than in a marketing sense for people with very little imagination.

During questions, @briankelly asked if any of the reported dabblings/demos (and there are several working demo) were just OUseful experiments or whether they could in principle be adopted within the OU, or even more widely across HE. The answers are ‘yes’ and ‘yes’ but in reality ‘yes’ and ‘no’. I haven’t even been able to get round to writing up (or persuading someone else to write up) any of my dabblings as ‘proper’ research, let alone fight the interminable rounds of lobbying and stakeholder acquisition it takes to get anything adopted as a rolled out as adopted innovation. If any of the ideas were/are useful, they’re Googleable and folk are free to run with them…but because they had no big budget holding champion associated with their creation, and hence no stake (even defensively) in seeing some sort of use from them, they unlikely to register anywhere.

Written by Tony Hirst

November 2, 2012 at 6:32 pm

Generating OpenLearn Navigation Mindmaps Automagically

I’ve posted before about using mindmaps as a navigation surface for course materials, or as way of bootstrapping the generation of user annotatable mindmaps around course topics or study weeks. The OU’s XML document format that underpins OU course materials, including the free course units that appear on OpenLearn, makes for easy automated generation of secondary publication products.

So here’s the next step in my exploration of this idea, a data sketch that generates a Freemind .mm format mindmap file for a range of OpenLearn offerings using metadata puled into Scraperwiki. The file can be downloaded to your desktop (save it with a .mm suffix), and then opened – and annotated – within Freemind.

You can find the code here: OpenLearn mindmaps.

By default, the mindmap will describe the learning outcomes associated with each course unit published on the Open University OpenLearn learning zone site.

By hacking the view URL, other mindmaps are possible. For example, we ca make the following additions to the actual mindmap file URL (reached by opening the Scraperwiki view) as follows:

  • ?unit=UNITCODE, where UNITCODE= something like T180_5 or K100_2 and you will get a view over section headings and learning outcomes that appear in the corresponding course unit.
  • ?unitset=UNITSET where UNITSET= something like T180 or K100 – ie the parent course code from which a specific unit was derived. This view will give a map showing headings and Learning Outcomes for all the units derived from a given UNITSET/course code.
  • ?keywordsearch=KEYWORD where KEYWORD= something like: physics This will identify all unit codes marked up with the keyword in the RSS version of the unit and generate a map showing headings and Learning Outcomes for all the units associated with the keyword. (This view is still a little buggy…)

In the first iteration, I haven’t added links to actual course units, so the mindmap doesn’t yet act as a clickable navigation surface, but that it is on the timeline…

It’s also worth noting that there is a flash browser available for simple Freemind mindmaps, which means we could have an online, in-browser service that displays the mindmap as such. (I seem to have a few permissions problems with getting new files onto ouseful.open.ac.uk at the moment – Mac side, I think? – so I haven’t yet been able to demo this. I suspect that browser security policies will require the .mm file to be served from the same server as the flash component, which means a proxy will be required if the data file is pulled from the Scraperwiki view.)

What would be really nice, of course, would be an HTML5 route to rendering a JSONified version of the .mm XML format… (I’m not sure how straightforward it would be to port the original Freemind flash browser Actionscript source code?)

Written by Tony Hirst

May 4, 2012 at 2:14 pm

Posted in Open Content, OU2.0

Tagged with , ,

Deconstructing OpenLearn Units – Glossary Items, Learning Outcomes and Image Search

It turns out that part of the grief I encountered here in trying to access OpenLearn XML content was easily resolved (check the comments: mechanise did the trick…), though I’ve still to try to sort out a workaround for accessing OpenLearn images (a problem described here)), but at least now I have another stepping stone: a database of some deconstructed OpenLearn content.

Using Scraperwiki to pull down and parse the OpenLearn XML files, I’ve created some database tables that contain the following elements scraped from across the OpenLearn units by this OpenLearn XML Processor:

  • glossary items;
  • learning objectives;
  • figure captions and descriptions.

You can download CSV data files corresponding to the tables, or the whole SQLite database. (Note that there is also an “errors” table that identifies units that threw an error when I tried to grab, or parse, the OpenLearn XML.)

Unfortunately, I haven’t had a chance yet to pop up a view over the data (I tried, briefly, but today was another of those days where something that’s probably very simple and obvious prevented me from getting the code I wanted to write working; if anyone has an example Scraperwiki view that chucks data into a sortable HTML table or a Simile Exhibit searchable table, please post a link below; or even better, add a view to the scraper:-)

So in the meantime, if ypu want to have a play, you need to make use of the Scraperwiki API wizard.

Here are some example queries:

  • a search for figure descriptions containing the word “communication” – select * from `figures` where desc like ‘%communication%’: try it
  • a search over learning outcomes that include the phrase how to followed at some point by the word dataselect * from `learningoutcomes` where lo like ‘%how to%data%’: try it
  • a search of glossary items for glossary terms that contain the word “period” or a definition that contains the word “ancient” – select * from `glossary` where definition like ‘%ancient%’ or term like ‘%period%’: try it
  • find figures with empty captions – select * from `figures` where caption==”: try it

I’ll try to add some more examples when I get a chance, as well as knocking up a more friendly search interface. Unless you want to try…?!;-)

Written by Tony Hirst

March 15, 2012 at 10:59 am

Do We Need an OpenLearn Content Liberation Front?

For me, one of the defining attributes of openness relates to accessibility of the machine kind: if I can’t write a script to handle the repetitive stuff for me, or can’t automate the embedding of image and/or video resources, then whatever the content is, it’s not open enough in a practical sense for me to do what I want with it.

So here’s an, erm, how can I put this politely, little niggle I have with OpenLearn XML. (For those of you not keeping up, one of the many OpenLearn sites is the OU’s open course materials site; the materials published on the site as course unit contentful HTML pages are also available as structured XML documents. (When I say “structured”, I mean that certain elements of the materials are marked up in a semantically meaningful way; lots of elements aren’t, but we have to start somewhere ;-))

The context is this: following on from my presentation on Making More of Structured Course Materials at the eSTeEM conference last week, I left a chat with Jonathan Fine with the intention of seeing what sorts of secondary product I could easily generate from the OpenLearn content. I’m in the middle of building a scraper and structured content extractor at the moment, grabbing things like learning outcomes, glossary items, references and images, but I almost immediately hit a couple of problems, first with actually locating the OU XML docs, and secondly locating the images…

Getting hold of a machine readable list of OpenLearn units is easy enough via the OpenLearn OPML feed (much easier to work with than the “all units” HTML index page). Units are organised by topic and are listed using the following format:

<outline type="rss" text="Unit content for Water use and the water cycle" htmlUrl="http://openlearn.open.ac.uk/course/view.php?name=S278_12" xmlUrl="http://openlearn.open.ac.uk/rss/file.php/stdfeed/4307/S278_12_rss.xml"/>

URLs of the form http://openlearn.open.ac.uk/course/view.php?name=S278_12 link to a ‘homepage” for each unit, which then links to the first page of actual content, content which is also available in XML form. The content page URLs have the form http://openlearn.open.ac.uk/mod/oucontent/view.php?id=398820&direct=1, where the ID is one-one uniquely mapped to the course name identifier. The XML version of the page can then be accessed by changing direct=1 in the URL to content=1. Only, we don’t know the mapping from course unit name to page id. The easiest way I’ve found of doing that is to load in the RSS feed for each unit and grab the first link URL, which points the first HTML content page view of the unit.

I’ve popped a scraper up on Scraperwiki to build the lookup for XML URLs for OpenLearn units – OpenLearn XML Processor:

import scraperwiki

from lxml import etree

#===
#via http://stackoverflow.com/questions/5757201/help-or-advice-me-get-started-with-lxml/5899005#5899005
def flatten(el):           
    result = [ (el.text or "") ]
    for sel in el:
        result.append(flatten(sel))
        result.append(sel.tail or "")
    return "".join(result)
#===

def getcontenturl(srcUrl):
    rss= etree.parse(srcUrl)
    rssroot=rss.getroot()
    try:
        contenturl= flatten(rssroot.find('./channel/item/link'))
    except:
        contenturl=''
    return contenturl

def getUnitLocations():
    #The OPML file lists all OpenLearn units by topic area
    srcUrl='http://openlearn.open.ac.uk/rss/file.php/stdfeed/1/full_opml.xml'
    tree = etree.parse(srcUrl)
    root = tree.getroot()
    topics=root.findall('.//body/outline')
    #Handle each topic area separately?
    for topic in topics:
        tt = topic.get('text')
        print tt
        for item in topic.findall('./outline'):
            it=item.get('text')
            if it.startswith('Unit content for'):
                it=it.replace('Unit content for','')
                url=item.get('htmlUrl')
                rssurl=item.get('xmlUrl')
                ccu=url.split('=')[1]
                cctmp=ccu.split('_')
                cc=cctmp[0]
                if len(cctmp)>1: ccpart=cctmp[1]
                else: ccpart=1
                slug=rssurl.replace('http://openlearn.open.ac.uk/rss/file.php/stdfeed/','')
                slug=slug.split('/')[0]
                contenturl=getcontenturl(rssurl)
                print tt,it,slug,ccu,cc,ccpart,url,contenturl
                scraperwiki.sqlite.save(unique_keys=['ccu'], table_name='unitsHome', data={'ccu':ccu, 'uname':it,'topic':tt,'slug':slug,'cc':cc,'ccpart':ccpart,'url':url,'rssurl':rssurl,'ccurl':contenturl})

getUnitLocations()

The next step in the plan (because I usually do have a plan; it’s hard to play effectively without some sort of direction in mind…) as far as images goes was to grab the figure elements out of the XML documents and generate an image gallery that allows you to search through OpenLearn images by title/caption and/or description, and preview them. Getting the caption and description from the XML is easy enough, but getting the image URLs is not

Here’s an example of a figure element from an OpenLearn XML document:

<Figure id="fig001">
<Image src="\\DCTM_FSS\content\Teaching and curriculum\Modules\Shared Resources\OpenLearn\S278_5\1.0\s278_5_f001hi.jpg" height="" webthumbnail="false" x_imagesrc="s278_5_f001hi.jpg" x_imagewidth="478" x_imageheight="522"/>
<Caption>Figure 1 The geothermal gradient beneath a continent, showing how temperature increases more rapidly with depth in the lithosphere than it does in the deep mantle.</Caption>
<Alternative>Figure 1</Alternative>
<Description>Figure 1</Description>
</Figure>

Looking at the HTML page for the corresponding unit on OpenLearn, we see it points to the image resource file at http://openlearn.open.ac.uk/file.php/4178/!via/oucontent/course/476/s278_5_f001hi.jpg:

So how can we generate that image URL from the resource link in the XML document? The filename is the same, but how can we generate what are presumably contextually relevant path elements: http://openlearn.open.ac.uk/file.php/4178/!via/oucontent/course/476/

If we look at the OpenLearn OPML file that lists all current OpenLearn units, we can find the first identifier in the path to the RSS file:

<outline type="rss" text="Unit content for Energy resources: Geothermal energy" htmlUrl="http://openlearn.open.ac.uk/course/view.php?name=S278_5" xmlUrl="http://openlearn.open.ac.uk/rss/file.php/stdfeed/4178/S278_5_rss.xml"/>

But I can’t seem to find a crib for the second identifier – 476 – anywhere? Which means I can’t mechanise the creation of links to actually OpenLearn image assets from the XML source. Also note that there are no credits, acknowledgements or license conditions associated with the image contained within the figure description. Which also makes it hard to reuse the image in a legal, rights recognising sense.

[Doh - I can surely just look at URL for an image in an OpenLearn unit RSS feed and pick the path up from there, can't I? Only I can't, because the image links in the RSS feeds are: a) relative links, without path information, and b) broken as a result...]

Reusing images on the basis of the OpenLearn XML “sourcecode” document is therefore: NOT OBVIOUSLY POSSIBLE.

What this suggests to me is that if you release “source code” documents, they may actually need some processing in terms of asset resolution that generates publicly resolvable locators to assets if they are encoded within the source code document as “private” assets/non-resolvable identifiers.

Where necessary, acknowledgements/credits are provided in the backmatter using elements of the form:

<Paragraph>Figure 7 Willes-Richards, J., et al. (1990) ; HDR Resource/Economics’ in Baria, R. (ed.) <i>Hot Dry Rock Geothermal Energy</i>, Copyright CSM Associates Limited</Paragraph>

Whilst OU-XML does support the ability to make a meaningful link to a resource within the XML document, using an element of the form:

<CrossRef idref="fig007">Figure 7</CrossRef>

(which presumably uses the Alternative label as the cross-referenced identifier, although not the figure element id (eg fig007) which is presumably unique within any particular XML document?), this identifier is not used to link the informally stated figure credit back to the uniquely identified figure element?

If the same image asset is used in several course units, there is presumably no way of telling from the element data (or even, necessarily, the credit data?) whether the images are in fact one and the same. That is, we can’t audit the OpenLearn materials in a text mechanised way to see whether or not particular images are reused across two or more OpenLearn units.

Just in passing, it’s maybe also worth noting that in the above case at least, a description for the image is missing. In actual OU course materials, the description element is used to capture a textual description of the image that explicates the image in the context of the surrounding text. This represents a partial fulfilment of accessibility requirements surrounding images and represents, even if not best, at least effective practice.

Where else might content need liberating within OpenLearn content? At the end of the course unit XML documents, in the “backmatter” element, there is often a list of references. References have the form:

<Reference>Sheldon, P. (2005) Earth’s Physical Resources: An Introduction (Book 1 of S278 Earth’s Physical Resources: Origin, Use and Environmental Impact), The Open University, Milton Keynes</Reference>

Hmmm… no structure there… so how easy would it be to reliably generate a link to an authoritative record for that item? (Note that other records occasionally use presentational markup such as italics (or emphasis) tags to presentationally style certain parts of some references (confusing presentation with semantics…).)

Finally, just a quick note on why I’m blogging this publicly rather than raising it, erm, quietly within the OU. My reasoning is similar to the reasoning we use when we tell students to not be afraid of asking questions, because it’s likely that others will also have the same question… I’m asking a question about the structure of an open educational resource, because I don’t quite understand it; by asking the question in public, it may be the case that others can use the same questioning strategy to review the way they present their materials, so when I find those, I don’t have to ask similar sorts of question again;-)

PS sort of related to this, see TechDis’ Terry McAndrew’s Accessible courses need and accessibilty-friendly schema standard.

PPS see also another take on ways of trying to reduce cognitive waste – Joss Winn’s latest bid in progress, which will examine how the OAuth 2.0 specification can be integrated into a single sign on environment alongside Microsoft’s Unified Access Gateway. If that’s an issue or matter of interest in your institution, why not fork the bid and work it up yourself, or maybe even fork it and contribute elements back?;-) (Hmm, if several institutions submitted what was essentially the same bid from multiple institutions, how would they cope during the marking process?!;-)

Written by Tony Hirst

March 13, 2012 at 1:10 pm

eSTeEM Conference Presentation – Making More of Structured Course Materials

A copy of the presentation I gave at the OU-eSTeEM conference (no event URL?) on generating custom course search engines and mining OU XML documents to generate course mindmaps (Making More of Structured Documents presentation; delicious stack/bookmark list of related resources):

Chatting to Jonathan Fine after the event, he gave me the phrase secondary products to describe things like course mindmaps that can be generated from XML source files of OU course materials. From what I can tell, there isn’t much if any work going on in the way of finding novel ways of exploiting the structure of OU structured course materials, other than using them simply as a way of generating different presentational views of the course materials as a whole (that is, HTML versions, maybe mobile friendly versions, PDF versions). (If that’s not the case, please feel free to put me right in the comments:-)

One thing Jonathan has been scouring the documents for is evidence of mathematical content across the courses; he also mentioned a couple of ideas relating to access audits over the content itself, such as extracting figure headings, or image captions. (This reminded me of the OpenLearn XML processor (and redux) I first played with 4 years ago (sigh… and nothing’s changed… sigh….), which stripped assets by type from the first generation of OU XML docs). So on my to do list is to have a deeper look at the structure of OU XML, have a peek at what sorts of things might meaningfully (and easily;-) extracted, and figure out two or three secondary products that can be generated as a result. Note that these products might be products for different audiences, at different times of the course lifecycle: tools for use by course team or LTS during production (such as accessibility checks), products to support maintenance (there is already a link checker, but maybe there is more that can be done here?), products for students (such as the mindmap), products for alumni, products for OpenLearn views over the content, products to support “learning analytics”, and so on. (If you have any ideas of what forms the secondary products might take, or what structures/elements/entities you’d like to see mined from OU XML, please let me know via the comments. For an example of an OU XML doc, see here.

Written by Tony Hirst

March 8, 2012 at 9:38 am

Search Engine Powered Courses…

How can we use customised search engines to support uncourses, or the course models used to support MOOC style offerings?

To set the scene, here’s what Stephen Downes wrote recently on the topic of How to partcipate in a MOOC:

You will notice quickly that there is far too much information being posted in the course for any one person to consume. We tried to start slowly with just a few resources, but it quickly turns into a deluge.

You will be provided with summaries and links to dozens, maybe hundreds, maybe even thousands of web posts, articles from journals and magazines, videos and lectures, audio recordings, live online sessions, discussion groups, and more. Very quickly, you may feel overwhelmed.

Don’t let it intimidate you. Think of it as being like a grocery store or marketplace. Nobody is expected to sample and try everything. Rather, the purpose is to provide a wide selection to allow you to pick and choose what’s of interest to you.

This is an important part of the connectivist model being used in this course. The idea is that there is no one central curriculum that every person follows. The learning takes place through the interaction with resources and course participants, not through memorizing content. By selecting your own materials, you create your own unique perspective on the subject matter.

It is the interaction between these unique perspectives that makes a connectivist course interesting. Each person brings something new to the conversation. So you learn by interacting rather than by mertely consuming.

When I put together the the OU course T151, the original vision revolved around a couple of principles:

1) the course would be built in part around materials produced in public as part of the Digital Worlds uncourse;

2) each week’s offering would follow a similar model: one or two topic explorations, plus an activity and forum discussion time.

In addition, the topic explorations would have a standard format: scene setting, and maybe a teaser question with answer reveal or call to action in the forums; a set of topic exploration questions to frame the topic exploration; a set of resources related to the topic at hand, organised by type (academic readings (via a libezproxy link for subscription content so no downstream logins are required to access the content), Digital Worlds resources, weblinks (industry or well informed blogs, news sites etc), audio and video resources); and a reflective essay by the instructor exploring some of the themes raised in the questions and referring to some of the resources. The aim of the reflective essay was to model the sort of exploration or investigation the student might engage in.

(I’d probably just have a mixed bag of resources listed now, along with a faceting option to focus in on readings, videos, etc.)

The idea behind designing the course in this way was that it would be componentised as much as possible, to allow flexibility in swapping resources or even topics in and out, as well as (though we never managed this), allowing the freedom to study the topics in an arbitrary order. Note: I realised today that to make the materials more easily maintainable, a set of ‘Recent links’ might be identified that weren’t referred to in the ‘My Reflections’ response. That is, they could be completely free standing, and would have no side effects if replaced.

As far as the provision of linked resources went, the original model was that the links should be fed into the course materials from an instructor maintained bookmark collection (for an early take on this, see Managing Bookmarks, with a proof of concept demo at CourseLinks Demo (Hmmm, everything except the dynamic link injection appears to have rotted:-().

The design of the questions/resources page was intended to have the scoping questions at the top of the page, and then the suggested resources presented in a style reminiscent of a search engine results listing, the idea being that we would present the students with too many resources for them to comfortably read in the allocated time, so that they would have to explore the resources from their own perspective (eg given their current level of understanding/knowledge, their personal interests, and so on). In one of my more radical moments, I suggested that the resources would actually be pulled in from a curated/custom search engine ‘live’, according to search terms specially selected around the current topic and framing questions, but I was overruled on that. However, the course does have a Google custom search engine associated with it which searches over materials that are linked to from the course.

So that’s the context…

Where I’m at now is pondering how we can use an enhanced custom search engine as a delivery platform for a resource based uncourse. So here’s my first thought: using a Google Custom Search Engine populated with curated resources in a particular area, can we use Google CSE Promotions to help scaffold a topic exploration?

Here’s my first promotions file:

<Promotions>
   <Promotion id="t151_1a" 
        queries="topic 1a, Topic 1A, topic exploration 1a, topic exploration 1A, topic 1A, what is a game, game definition" 
        title="T151 Topic Exploration 1A - So what is a game?" 
        url="http://digitalworlds.wordpress.com/2008/03/05/so-what-is-a-game/"
        description="The aim of this topic is to think about what makes a game a game. Spend a minute or two to come up with your own definition. If you're stuck, read through the Digital Worlds post 'So what is a game?'"
        image_url="http://kmi.open.ac.uk/images/ou-logo.gif" />
</Promotions>

It’s running on the Digital Worlds Search Engine, so if you want to try it out, try entering the search phrase what is a game or game definition.

T151 CSE promotion - game definition

(This example suggests to me that it would also make sense to use result boosting to boost the key readings/suggested resources I proposed in the topic materials so that they appear nearer the top of the results (that’ll be the focus of a future post;-))

The promotion displays at the top of the results listing if the specified queries match the search terms the user enters. My initial feeling is that to bootstrap the process, we need to handle:

- queries that allow a user to call on a starting point for a topic exploration by specifically identifying that topic;
– “naive queries”: one reason for using the resource-search model is to try to help students develop effective information skills relating to search. Promotions (and result boosting) allow us to pick up on anticipated naive queries (or popular queries identified from search logs), and suggest a starting point for a sensible way in to the topic. Alternatively, they could be used to offer suggestions for improved or refined searches, or search strategy hints. (I’m reminded of Dave Pattern’s work with guided searches/keyword refinements in the University of Huddersfield Library catalogue in this context).

Here’s another example using the same promotion, but on a different search term:

T151 CSE - topic 1a

Of course, we could also start to turn the search engine into something like an adventure game engine. So for example, if we type: start or about, we might get something like:

T151 CSE - start

(The link I associated with start should really point to the course introduction page in the VLE…)

We can also use the search context to provide pastoral or study skills support:

T151 CSE - pastoral

These sort of promotions/enhancements might be produced centrally and rolled out across course search engines, leaving the course and discipline related customisations to the course team and associated subject librarians.

Just a final note: ignoring resource limitations on Google CSEs for a moment, we might imagine the following scenarios for their role out:

1) course wide: bespoke CSEs are commissioned for each course, although they may be supplemented by generic enhancements (eg relating to study skills);

2) qualification based: the CSE is defined at the qualification level, and students call on particular course enhancements by prefacing the search with the course code; it might be that students also see a personalised view of the qualification CSE that is tuned to their current year of study.

3) university wide: the CSE is defined at the university level, and students students call on particular course or qualification level enhancements by prefacing the search with the course or qualification code.

Written by Tony Hirst

September 15, 2011 at 2:03 pm

Tweaking Ranking Factors in the Course Detective Custom Search Engine

This is a note-to-self as much as anything, relating to the Course Detective custom search engine that searches over UK HE course prospectus web pages about the extent to which we might be able to use data such as the student satisfaction survey results (as made available via Unistats) to boost search results around particular subjects in line with student satisfaction ratings, or employment prospects, for particular universities?

It’s possible to tweak rankings in Google CSEs in a variety of ways. On the one hand, we can BOOST (improve the ranking), FILTER (limit results to members of a given set) or ELIMINATE (exclude) sites appearing in the search results listing. In the simplest case, we assign a BOOST, FILTER or ELIMINATE weight to a label, and then apply labels to annotations so that they benefit from the corresponding customisation. We can further refine the effect of the modification by applying a score to each annotation. The product of score and weight values determines the overall ranking modification that is applied to each result for a label applied to an annotation.

So here’s what I’m thinking:

- define labels for things like achievement or satisfaction that apply a boost to a result;
– allow uses to apply a label to a search;
– for each university annotation (that is, the listing that identifies the path to the pages for a particular university’s online prospectus), add a label with a score modifier determined by the achievement or satisfaction value, for example, for that institution;
– for refinement labels that tweak search rankings within a particular subject area, define labels corresponding to those subject areas and apply score modifiers to each institution based on, for example, the satisfaction level with that subject area. (Note: I’m not sure if the same path can have several different annotations provided to it with different scores?

For example, an annotation file typically contains a fragment that looks like:

<Annotations>
  <Annotation about="webcast.berkeley.edu/*" score="1">
    <Label name="university_boost_highest"/>
    <Label name="lectures"/>
  </Annotation>

  <Annotation about="www.youtube.com/ucberkeley/*" score="1">
    <Label name="university_boost_highest"/>
    <Label name="videos_boost_mid"/>
    <Label name="lectures"/>
  </Annotation>
</Annotations>

I don’t know if this would work:

<Annotations>
  <Annotation about="example.com/prospectus/*" score="1">
    <Label name="chemistry"/>
  </Annotation>
<Annotations>
  <Annotation about="example.com/prospectus/*" score="0.5">
    <Label name="physics"/>
  </Annotation>
</Annotations>

That said, if the URLs are nicely structured, we might be able to do something like:

<Annotations>
  <Annotation about="example.com/prospectus/chemistry/*" score="1">
    <Label name="chemistry"/>
  </Annotation>
<Annotations>
  <Annotation about="example.com/prospectus/physics/*" score="0.5">
    <Label name="physics"/>
  </Annotation>
</Annotations>

albeit at the cost of having to do a lot more work in terms of identifying appropriate URI paths.

I also need to start thinking a bit more about how to apply refinements and ranking adjustments in course based CSEs.

Written by Tony Hirst

August 5, 2011 at 2:51 pm

Posted in Anything you want, Search

Tagged with ,

Follow

Get every new post delivered to your Inbox.

Join 799 other followers