« FRBR & "Time-Based Media", Part 4: Alternate Forms & Supplementary Materials | Main | Great expectations? »

June 10, 2008

Talis Xiphos Research Day

This blog entry was used to host a live blog report from the Talis Xiphos Research Day, held at the Talis offices outside Birmingham on 10 June 2008.

Note that this page, and the live blog it contains, has been edited after the event to correct minor typos and so on.  However, no substantial changes have been made.

The day's agenda was:

09:00 Coffee & Registration
09:30 Welcome & Introduction
09:45 Peter Murray-Rust, University of Cambridge - Data-Driven research.
10:30 Andy Powell, Eduserv - Web 2.0 and repositories - have we got our repository architecture right?
11:15 Coffee
11:30 Carsten Ulrich, Shanghai Jiao Tong University - Why Web 2.0 is Good for Learning and for Research: Principles and Prototypes.
12:15 Alan Masson, Senior Lecturer in Learning Technologies, University of Ulster - Formalising the informal - using a Hybrid Learning Model to Describe Learning Practices.
13:00 Lunch
13:45 Chris Clarke, Talis - Project Xulu - Creating a Social Network from a Web of Scholarly Data.
14.15 Alan Trow-Poole and Ian Corns, Talis - Project Zephyr: Letting Students Weave Their Own Path.
14:45 Attendee discussion and feedback
16.00 Close

As you may note from the live blog transcript, there was a small modification to the afternoon agenda, adding Nadeem Shabir (Talis) at the end of the programme.

Clearly, I was unable to live blog myself.  Rob Styles and Nadeem Shabir added comments to the live blogging system as my talk progressed.  Unfortunately, comments in the Coveritlive system that I was using need to be approved by the live blog author in real time and I was not available to do so, meaning that their comments only became visible after I had finished my talk and had started live blogging the next speaker.  Apologies that things go slightly out of order at that point.

Overall, the day was very enjoyable and interesting, with talks covering a range of topics relevant to learning and research.

Talis Xiphos Research Day (06/10/2008)
Powered by: CoveritLive
9:36
Andy Powell -  ok, Peter Murray Rust (PMR) is just starting...
9:37
[Comment From AdrianStevenson]
Don't suppose there's any audio feed?
9:38
Andy Powell -  @ade no, sorry, no audio - i may try setting something up later, but too rushed now with my talk coming up next
9:40
Andy Powell -  PMR saying thast publishers are primarily about preventing access to content - then asking if there are any publishers in the room - which there are! lol
9:40
Adrian Stevenson -  @AndyPowell OK. Shame I didn't spot this event. as would have attended An audio recording at least would be great. Slides anywhere?
9:40
Andy Powell -  my slides are at http://tinyurl.com/4fehq8
9:40
Adrian Stevenson -  cheers
9:40
Andy Powell -  not sure about others
9:41
Andy Powell -  PMR not using slides as such
9:42
Andy Powell -  PMR currently outlining topic of his talk - open data, repositories, etheses, semantic data, science commons, ...
9:42
Andy Powell -  openNoteBook science
9:44
Andy Powell -  showing graph of atmospheric carbon dioxide
9:45
Andy Powell -  arguing that ability to share digital copies of data and analysis (a graph in this case) very important
9:46
Andy Powell -  scientific publication = discourse (human readable), embedded facts, etc. but using of PDF prevents machine re-use
9:47
Adrian Stevenson -  Cheers for this micro blogging btw Andy. Don't feel obliged to do it all day!
9:47
Andy Powell -  i find it helpful for me :-)
9:47
[Comment From Paul Miller]
Presentations should go up at www.talis.com/xiphos/events shortly.
9:49
Andy Powell -  in chemistry, some data available thru some chimcal databases (repositories) but embedded chemical formula, etc. only available in human-readable form
9:49
Adrian Stevenson -  Ta Paul
9:50
Andy Powell -  patent office is looking at sharing semantic data
9:50
Andy Powell -  this disrupts gatekeepers who currently re-purpose non-semantic data for industry
9:51
Andy Powell -  gatekeepers currently quite powerful
9:52
Andy Powell -  talking hangovers from paper age of publishing - from victorian age - still affecting the way publishing happens today - e.g publishers asking for graphs to be removed from papers to save "space"
9:53
Andy Powell -  talking about the importance of publishing data as well as results - so that results can be verified, re-tested, etc.
9:53
Adrian Stevenson -  Anybody there able to cover Andy's talk? Unless he's going to do both ..
9:54
Andy Powell -  now showing 'real' scientists at work - journal of visualized experiments www.jove.com
9:55
Andy Powell -  reports of expeiments using video
9:55
Andy Powell -  and text explanations
9:57
Andy Powell -  noting importance of publishing science in sufficient detail that it can be repeated
9:57
Andy Powell -  big science epitomises the data deluge and data is well supported
9:58
Andy Powell -  long-tail science also generates lots of data
9:59
Andy Powell -  err, no, i aint going to do both - but i'm just saying the same old stuff :-) repositpories, tada yada, web 2.0 yada yada, semantic web yada yada - that kind of thing
10:00
Andy Powell -  most science done in the long tail - small lab work and so on
10:00
Andy Powell -  how do we deal with the long tail of science - not currently well catered for
10:00
Andy Powell -  repositories don't currently meet the needs of long tail science
10:01
Andy Powell -  paul might cover my bit - paul, you can comment without moderation - hint hint
10:02
Andy Powell -  domain repositories currently cater better for long tail science - you don't use your institutional repository
10:02
Andy Powell -  30% of scientists have lost their data at some point
10:03
Andy Powell -  showing a typical thesis
10:03
Andy Powell -  graphs, molecules, equations, ...
10:04
Andy Powell -  PDF == reading with boxing gloves on
10:04
Andy Powell -  can't get at the interesting data that is embedded in the document
10:04
Andy Powell -  showing robot (called oscar) that reads word files and extracts interesting data
10:05
Andy Powell -  xml exposed by word much better than pdf in terms of processability
10:06
Andy Powell -  table constructed automatically from text of document
10:07
Andy Powell -  can process large part of the chemical literature using this kind of technology

10:07
Andy Powell -  but... copyright (to publishers) prevents re-use
10:08
[Comment From PeteJ]
This sounds similar content to PM-R's OR08 keynote session. Not that that's bad thing. Just sayin'.
10:09
Andy Powell -  demoing another tool called CrystalEye - wwmm.ch.cam.ac.uk/crystaleye (I think)
10:10
Andy Powell -  grabs ToC for laterst issue of journal and then draws molecules that are not in the original publication based on info extracted
10:10
PeteJ -  Memories of fierce MSWord v PDF arguments between text miners & preservationists @ OR08
10:11
Andy Powell -  PMRs robots could go thru whole chemical literature and open it up - but strong business/publisher lobby prevents it
10:11
Andy Powell -  now showing what can be done with theses
10:13
Andy Powell -  chemistry "not much fun to read in bed"! lol
10:13
Andy Powell -  you heard it here first
10:14
Andy Powell -  showing more automatic analysis of research data (based on RDF data I think)
10:15
Andy Powell -  i'm getting a bit lost here - talking about chemical data which i don't understand - think the underlying message is that a picture is worth a thousand words - but the way we share words prevent people from drawing pictures easily
10:16
Andy Powell -  ah... now onto repositories
10:16
Andy Powell -  asking about sourceforge and eclipse
10:17
Andy Powell -  sourceforge is a repository - for managing computer code
10:17
Andy Powell -  collaborative environment
10:17
Andy Powell -  i.e. it supports social interactions
10:18
[Comment From Guest]
@andypowell - reading chemistry in bed probably wasn't what was making the floorboards squeak in your hotel...
10:18
[Comment From Silversprite]
Bloated PDF can be nightmare for researchers handicapped by tide-dependent broadband :-(
10:19
Andy Powell -  sorry - i'm being slow to approve comments!   concentrating on presentation too hard
10:20
Andy Powell -  talking about bioclipse - open source tool
10:23
Andy Powell -  now moving on to science commons
10:23
Andy Powell -  outgrowth of creative commons
10:24
[Comment From Paul Miller]
Wondering if Owen can take over when Andy speaks, next?
10:24
Andy Powell -  possibly... but think i'm going to present from my own machine
10:25
Andy Powell -  need to protect our data (as being open) before others come along and "steal" it
10:25
Andy Powell -  explicitly flag data as being "open"
10:25
Andy Powell -  talking about publishers agin - it's NOT their data!
10:29
Andy Powell -  in Q and A
10:29
Andy Powell -  I'm up next... so live blogging will stop for next 45 minutes
10:30
[Comment From Chris Keene]
Thanks Andy, hopefully someone will blog a summary of your presentation to go along with the slides.
10:33
Adrian Stevenson -  Don't forget to hit the Audacity record button Andy
11:23
Andy Powell -  ok, i'm back now... but we are going into coffe so there'll be a brief pause
11:38
Andy Powell -  right... next session about to start
11:38
Andy Powell -  Carsten Ulrich, Shanghai Jiao Tong University - Why Web 2.0 is Good for Learning and for Research: Principles and Prototypes
11:39
[Comment From Nadeem Shabir]
Andy Powell starts his talk ... on Web 2.0 and Institutional Repositories
11:39
[Comment From Nadeem Shabir]
andy talking about how we need to get the achitecture for repositories right
11:39
[Comment From Rob Styles]
"PDF is a cul-de-sac" nice quote.
11:39
[Comment From Rob Styles]
"we tend to focus on service-oriented approaches" that is we focus on services on the content, rather than the content itself
11:39
[Comment From Rob Styles]
Flickr, YouTube et al - successful repositiories, based on the social activity surrounding the content
11:39
[Comment From Rob Styles]
insitutional repositiories don't match the social networks of researchers, which are subject-based, cross-institutional and global.
11:39
[Comment From Rob Styles]
Andy suggesting that global subject-centric repositories as a possible solution
11:39
[Comment From Rob Styles]
nice quote "Thou Shalt Deposit" to force content into repositories that would otherwise remain empty
11:39
[Comment From Rob Styles]
arxiv.org as good example of how to do it, but started before we knew how to scale things like that
11:39
[Comment From Rob Styles]
Q: Why do blogs work when institutional repositories don't?
11:39
[Comment From Rob Styles]
very difficult to have a real conversation about what's best as many just want to get the next step - opening up the research using reporitories and OAI-PMH - done and embedded.
11:39
[Comment From Rob Styles]
slideshare shown as an example of a web2.0 repository
11:39
[Comment From Rob Styles]
Summary: Go Simple - RSS, Tagging, full-text indexing, microformats (maybe)
11:39
[Comment From Rob Styles]
Alternatively, we look to the semantic web and add real meaning
11:39
[Comment From Rob Styles]
references to FRBR, SWAP, ORE and DCAM
11:40
[Comment From Rob Styles]
Digital Photography and Flickr fundamentally changed the nature of photography - the growth of something different and new,
11:40
[Comment From Rob Styles]
whereas scholarly publication is taking what we have done on paper and replicate it on the web.
11:40
[Comment From Rob Styles]
Carsten: "ah, so repositories are about making content available on the web"
11:40
[Comment From Rob Styles]
"I didn't get what arxiv.org was for, but had no problem understanding slideshare"
11:40
[Comment From Rob Styles]
andy: complexity is because we live in this hybrid web/paper world
11:41
Andy Powell -  apologies to rob and nadeem who tried to cover my talk - but i hadn't approved them in advance so the rest of you didn't see it
11:41
Andy Powell -  wish coveritlive had a 'approve all' mode
11:44
Andy Powell -  Carsten talking about web2.0 and learning
11:44
Andy Powell -  web 2.0 as a research tool
11:45
Andy Powell -  going to cover 3 examples i think
11:46
Andy Powell -  describing pedagogy associated with traditional teaching methods - teacher imparting knowledge thru lectures - little opportunity for discussion by students
11:46
Adrian Stevenson -  Cheers to Rob  and Nadeem. Doesn't really matter wasn't live for us remoters.
11:46
Andy Powell -  arguing that LMS (learning management systems) are teacher centred
11:47
Andy Powell -  intelligent tuoring system try to model the knowledge of experts - e.g. in physics
11:48
Andy Powell -  based on cognitive learning theories
11:48
Andy Powell -  can recognise when topics have been understood by the student - but very expensive
11:48
Andy Powell -  still based on the idea that expert has knowledge to impart to others
11:49
Andy Powell -  citing stephen downes as example of theorists that have embraced power of web 2.0
11:50
Andy Powell -  quoting confucious: "tell me and i'll forgot - show me and i may remmber - involve me and i'll know it forever" (sorry paraphrasing)
11:50
Andy Powell -  suggesting that there has been little analysis of the pedagogical value of Web 2.0
11:51
Andy Powell -  what is web 2.0?   tim o'reilly and so on...
11:51
Andy Powell -  participation
11:51
Andy Powell -  important in learning as way for students to express themselves and as new tools for teaching
11:52
Andy Powell -  "participation" as in expressing themselves in pictures, videos, etc.
11:52
Andy Powell -  facilitates "constructivist" learning
11:53
Andy Powell -  @silversprite session will remain on efoundations
11:53
Andy Powell -  requires open approach
11:54
PeteJ -  Jumping back a bit to Andy's presentation... while SlideShare, Flickr et al do the cross-institutional/social bit, they do tend to divide the world up by resource type. My "scholarly works" include papers, videos, still images, audio etc w relationships between them that cut across resource type boundaries. (This video is delivery of those slides based on that paper etc).
11:54
Andy Powell -  web 2.0 increases audience - but needs to be exploited
11:54
Andy Powell -  @petej yes, agreed
11:55
Andy Powell -  web 2.0 - huge variety of data available - thru apis - often annotated - increasingly semantic linkedData
11:56
Andy Powell -  interesting for teaching - gives ability to re-combine vast array of material from real networks - real contexts
11:56
Andy Powell -  open linked data - big potential for learning
11:57
Andy Powell -  architecture of assembly - access via apis - functionality via widgets
11:57
Andy Powell -  in education can build prototypes very quickly - e.g. using yahoo pipes
11:58
[Comment From Silversprite]
When the day is done, is there a way of getting this whole "Cover It Live" session (not in PDF!)?
11:58
Andy Powell -  talking about personal learning environments
11:58
Andy Powell -  showing scott wilson's diagram
11:59
Andy Powell -  PLE for language learning built from iGoogle - drag and drop to build it
11:59
Andy Powell -  web 2.0 = perpetual beta
12:00
Andy Powell -  in education the improvements in functionality can be good but also distracting
12:00
Andy Powell -  e.g. need to adapt manuals, etc.
12:00
Andy Powell -  but developers open to feedback - which is good
12:00
Andy Powell -  e.g. asking twitter developers to open up channels explicitly for learning application
12:01
Andy Powell -  web 2.0 independent access to data - long tail - lightweight models
12:01
Andy Powell -  principles of web2.0 enable social and active learning
12:02
Andy Powell -  best used for t&l when you exploit these features = active role for teachers
12:03
Andy Powell -  web 2.0 for research - multitude of services with aerchitecture of assembly - easy to combine stuff - quick prototyping
12:03
Andy Powell -  good for assessing hypotheses
12:03
Andy Powell -  ok, 2 examples...
12:04
Andy Powell -  1st example: learning resource creation - intention not to overload lecturers
12:04
Andy Powell -  authoring learning resources is hard, time consuming and costly
12:04
Andy Powell -  how can we help them?
12:05
Andy Powell -  hypothesis that social bookmrking should help them
12:05
Andy Powell -  get lecturers to use del.icio.us to bookmark resources - but with predefined tags for concepts, subjects, instructional types and difficulty/level
12:07
Andy Powell -  used this to extend the LMS by embedding links from del.icio.us - very low cost to implement prototype
12:07
Andy Powell -  why? - because of the del.icio.us api
12:08
Andy Powell -  feedback from lecturers was +ve - lecturer suggested also allowing students to tag resources
12:08
Andy Powell -  but students don't tag resources if textbook is good enough - so plan to use this approach on course where textbook material isn't rich enough
12:09
Andy Powell -  2nd example: using microblogging for language learning
12:09
Andy Powell -  vocational learners shy, seldom active, limited time
12:09
Andy Powell -  goal was to provide practice possibilities
12:10
Andy Powell -  microblogging (twitter?) increasing sense of community
12:10
Andy Powell -  encoutraged participation by reducing transactional distance to teacher - quick and easy way of active participation
12:11
Andy Powell -  yes, using twitter for this
12:11
Andy Powell -  example shown is EfL courses
12:11
Andy Powell -  implemented by downloading all twitter contributions - grades based on number of contributions - not quality of contributions
12:12
Andy Powell -  made use of twitter api - but also needed screen scraping because of limitations in twitter api
12:13
Andy Powell -  98 out of 110 students participated - 5574 twitter updates over 7 weeks
12:13
Andy Powell -  teachers also contributed - ~3 updates per day
12:13
Andy Powell -  questionaire at end of experiment
12:14
Andy Powell -  most students liked it - only 5% of students anti the use of twitter on the course
12:14
Andy Powell -  "twitter is the same as the schoolyard"
12:14
Andy Powell -  50% students said that they communicated not just with each other but with other native english speakers thru twitter
12:15
Andy Powell -  but - correction of mistakes typically not done
12:15
Andy Powell -  lessons learned...
12:15
Andy Powell -  web 2.0 can stimulate learning and participation
12:15
Andy Powell -  use of twitter continued after course had finished
12:16
Andy Powell -  social dimension was very important
12:16
Andy Powell -  students encouraged each other to participate
12:16
Andy Powell -  though sometimes students reverted to native language
12:16
Andy Powell -  teacher as both moderator and participator
12:17
Andy Powell -  little use of mobile devices - too many updates from twitter
12:17
Andy Powell -  no students integrated twitter into their blogs (they weren't shown how to do this)
12:18
Andy Powell -  now going to talk about Totuba Toolkit - start-up company in Shanghai
12:18
Andy Powell -  seeking feedback about whether what they are doing sounds useful
12:19
Andy Powell -  toolkit for creating and storing "notes"
12:19
Andy Powell -  automated intelligent suggestions for realted resources
12:19
Andy Powell -  find stuff and bookmark it
12:20
Andy Powell -  add addional info and notes - what kind of resource it is
12:20
Andy Powell -  ability to export reference list of bookmarked resources
12:21
Andy Powell -  visualise what has been collected in different ways - e.g. as knowledge map
12:23
Andy Powell -  goal of totuba is to facilitate process of learning and research by removing unnecessary steps - automate integration work - make it easier to find associated resources and peers
12:24
Andy Powell -  lessons learned again...   about use of Web 2.0 for elearning generally
12:24
Andy Powell -  architecture of asembly good - prototyping good
12:24
Andy Powell -  but - reliance on third party api
12:25
Andy Powell -  yet another login - but e.g. open social graph, and openid should help with this
12:25
Andy Powell -  not all functionality available via the api
12:25
Andy Powell -  web 2.0 less suitable for "designed instruction"
12:25
Andy Powell -  good at building community
12:26
Andy Powell -  but have to be prepared for side-effects - e.g. in the 2nd example students started using twitter avatar image to share photos!
12:26
Andy Powell -  requires active teacher - stimulating use of tools
12:26
Andy Powell -  again - noting reliance on third-party tools
12:27
Andy Powell -  linked open data is still for experts
12:28
Andy Powellslideshare.net/ullrich for the slides
12:28
Andy Powell -  ok, now taking questions...
12:30
Andy Powell -  PMR: HE is about getting a degree - not about learning - web 2.0 great for learning, poor for assessment ??
12:31
Andy Powell -  Alan Masson, Senior Lecturer in Learning Technologies, University of Ulster - Formalising the informal - using a Hybrid Learning Model to Describe Learning Practices - up next
12:34
Andy Powell -  technology provides opportunities but need to work out how to enable teachers and learners to take advantage of it
12:34
Andy Powell -  going to be talking about Hybrid Learning Model (HLM) and implications of its use to help lecturers reflect on their teaching practice
12:35
Andy Powell -  facilitating "learner centred" reflective practice by teachers - but need to change teaching practice in order for this to happen - not easy to do
12:36
Andy Powell -  need to describe current practice - disseminate new practice - ensuring learner is "core"
12:36
Andy Powell -  developed "modeling framework" to achieve this
12:37
Andy Powell -  practitioners have a comfort zone - focus on content and assessment
12:38
Andy Powell -  learning design - i.e. IMS LD spec. - provides basis for work
12:38
Andy Powell -  structure within which content and assessment can be placed
12:38
Andy Powell -  formal schemas and vocabs
12:39
Andy Powell -  But LD not reflective in nature - it's about design - UI of tools not yet mature - beta interfaces
12:40
Andy Powell -  HLM brings together "8LEM" model (Uni of Liege) and "Closed set of learning verbs" (Sue Bennett, Uni of Wollongong)
12:40
Andy Powell -  focus on interactions between participants
12:41
Andy Powell -  describing 8LEM - see http://cetl.ulster.ac.uk/elearning/index.php?page=8LEM-1 for details
12:42
Andy Powell -  now looking at the learning verbs
12:42
Andy Powell -  need to see slides for this bit!
12:43
Andy Powell -  we have been given cards in our delegate packs apparently
12:44
Andy Powell -  have used to cards with lecturers - facilitated, informal context (to improve reflection) with model transcribed into relevant data grid
12:45
Andy Powell -  how are cards used??
12:45
Andy Powell -  teaching staff bring along a set of objectives
12:45
Andy Powell -  1:1 sessions lasting 45 mins to 1 hours
12:46
Andy Powell -  describing cards being used to facilite understanding about what lecturer is trying to achieve (in a specific lesson)
12:48
Andy Powell -  trying to captuire what the teacher thinks they are doing and what they expect the students to be doing
12:49
Andy Powell -  link these activities to resources and something else - urghh, i'm struggling here
12:50
Andy Powell -  HLM results in text based grid and animated activity plan - but staff also like to see a mindmap
12:51
Andy Powell -  showing a completed grid - presented as an animated walk-thru
12:51
Andy Powell -  walk-thru shows teacher's role, learner's role, and what will be learned
12:52
Andy Powell -  intention is to help students understand the process rather than just the outcomes
12:53
Andy Powell -  result is that lecturers are formalising processes that haven't be articulated before
12:53
Andy Powell -  they are creating artifacts that formalise what they do - but also that challenge what they are doing
12:54
Andy Powell -  i.e. that challenge their teaching values
12:55
Andy Powell -  this model potentially helps bridge divide between "wooly" teaching practice (what actually happens in the classroom) and highly formalised constructs such as IMS LD
12:56
Andy Powell -  benefit of the model is that small chunks of structured information provides very useful building blocks for teachers
12:58
Andy Powell -  practitioner feedback about developoment and use of the model +ve - e.g. "encouraged me to think about leaner's perspective rather than just focusing on the teacher"
1:00
Andy Powell -  describing evaluation by learners of the use of the model - intention was to help year 1 students to new learning environment of uni
1:02
Andy Powell -  student: "helped me to bring everything together and know what is expected of me"
1:03
Andy Powell -  verbs helped students understand what processes were expected of them
1:08
Andy Powell -  sorry... i'm still here... but struggling to get my head round some of this... my problem, not the speaker's
1:10
Andy Powell -  use cases in which this approach is expected to be relevent...
1:11
Andy Powell -  raising awareness of learner perspective in teaching and learning processes
1:11
Andy Powell -  reflecting on and reviewing current practice
1:11
Andy Powell -  planning and designing course materials
1:11
Andy Powell -  providing reference framework for course administration
1:12
Andy Powell -  assisting students to adapt to new learning situations
1:13
Andy Powell -  summary - light model - easy to capture stuff - focus on practice - focus on learner perspective - multiple use cases - +ve evaluations so far - model formally embedded into Uni of Ulster (thru staff induction) - formalising the informal
1:15
Andy Powell -  one of the drivers for this is need to bring more diverse range of students and address issues around retention
1:15
Andy Powell -  also trying to address cultural issues between student base (partic. new students) and older base of existing staff
1:17
Andy Powell -  it'll be lunch in a minute... back in a while
1:20
Andy Powell -  assessment of model follows: "usability, use, impact" track - project is in 1st year - so not into 'impact' phase yet
1:20 [Be Right Back Countdown] 30 minutes
2:00
Chris Keene -  thinking Chris Clarke should just be taking stage?
2:06
Andy Powell -  ok, i'm back from lunch - sorry we are running slightly late now
2:07
Andy Powell -  Chris Clarke, Talis - Project Xulu - Creating a Social Network from a Web of Scholarly Data
2:07
Andy Powell -  describing the early web
2:08
Andy Powell -  what we have today is a web of documents - millions of documents - but they are human oriented - machines can't understand them
2:09
Chris Keene -  yes, how dare you take lunch away from your laptop :)
2:09
Andy Powell -  looking at a 'simple' google query "how many people were evacuated during hurricane katrina"
2:09
Andy Powell -  Google gives a fairly decent answer - from wikipedia page
2:10
Andy Powell -  but - more by luck than judgement
2:11
Andy Powell -  one of the problems with the web is that the meaning of links in hidden - not machine-understandable - makes page rank less useful than it might be
2:11
Andy Powell -  semantics of links cannot be determined by machines
2:11
Andy Powell -  arguing that we need a machine-readable web
2:12
Andy Powell -  back to wikipedia page about katrina - it contains lots of assertions about facts and so on
2:12
Andy Powell -  DBpedia has derived 218 million assertions out of wikipedia
2:13
Andy Powellwww.powerset.com have built a user-experience out of the DBpedia data - which is openly available
2:13
Andy Powell -  gives much better result than that obtained by simple Google search
2:15
Andy Powellhttp://dbpedia.org/
2:15
Andy Powell -  just like web of documents, the web of data is distributed
2:15
Andy Powell -  lots of participants
2:16
Andy Powell -  describing what talis is - i'll spare you the details - mentioning the talis platform
2:16
Andy Powell -  "doing the heavy lifting for the semantic web"
2:16
Andy Powell -  chris going to talk about project xiphos
2:17
Andy Powell -  what can we (talis) do using a web of scholarly data?
2:17
Andy Powell -  given metadata for 500 articles by a friendly company - sorry, i missed the name of the company
2:18
Andy Powell -  developed visualisation of the relationships between those articles
2:19
Rob Styles -  The visualisation is Relation Browser by Moritz Stefaner
2:19
Andy Powell -  thanks rob
2:19
Andy Powell -  different colors representing different relationships between entities in the data
2:19
Andy Powell -  19800 distinct articles - mainly thru citations
2:20
Andy Powell -  21029 people
2:20
Andy Powell -  instigated a small talis project to investigate how this data can be made useful
2:20
Rob Styles -  4 people for one month, for all design and coding
2:21
Andy Powell -  developed a scholarly social network prototype
2:22
Andy Powell -  demoing from PoV of real female researcher who is listed in the sample graph
2:22
Andy Powell -  search for "flowers" (a person's name - the name of one of her collaborators)

2:23
Andy Powell -  results categorised into 4 tabs - things, people, subjects, collections
2:23
Andy Powell -  TP Flower (the person she is looking for) appears under 'people' tab
2:24
Andy Powell -  Xiphos can pre-build 'home' page for people based on knowledge in the sample graph
2:24
Andy Powell -  'home' page shows 'work', 'knows' and 'collections' tabs
2:25
Andy Powell -  the xiphos system then allows end-users to augment the computed information by hand - e.g. by clicking on an 'i know this person' link
2:26
Andy Powell -  system prompts for registration info, then tries to marry up name and other info against knwoledge in the graph
2:26
Andy Powell -  takes newly registered user to their new homepage - a rich page because of knowledge in the graph
2:27
Andy Powell -  though it might also contain some errors because of fuzziness in the graph
2:28
Andy Powell -  info includes the person's 'network' - 4 types of relationships - people you know, people you cite, people who cite you, people that you are watching
2:28
Andy Powell -  the last of these gives people a way of "watching" (i.e. tracking) someone, without indicating that you formally know them
2:29
Andy Powell -  "watching" someone is a one way relationship i.e. you know who you are watching but you don't know who is watching you
2:30
Andy Powell -  clicking on a 'work' (i.e a publication) takes you to a page for that work - overview, citations, cited by, collections
2:31
Andy Powell -  also offers a thumbnail preview of the document itself - but small enough not to break copyright (arguably)
2:31
Andy Powell -  navigate thru citations both inbound and outbound
2:32
Rob Styles -  The thumbnails in the prototype were generated with permission ;-)
2:32
Andy Powell -  collections give a way to organise stuff in ways that are relevant to the end user
2:33
Andy Powell -  collections can be watched
2:33
Andy Powell -  people can be members of collections i think
2:33
Andy Powell -  ??
2:33
Rob Styles -  did he mention you can add people to collections
2:33
Andy Powell -  not sure
2:34
Andy Powell -  can also organise stuff by events
2:35
Rob Styles -  well, people can be collaborators on the creation and management of a collection, they can also be an entry in a collection - as in a collection of people of interest
2:35
Andy Powell -  (events not implemented - just wire-framed currently)
2:35
Andy Powell -  also offers 'repository' functionality in the form of a 'Vault'
2:36
Andy Powell -  bit like sourceforge - but could offer view across distributed set of repositories
2:36
Andy Powell -  why a social network?
2:37
Andy Powell -  because it encourages users to clean/correct the data in ways that can't be done purely automatically
2:37
Andy Powell -  what else can be done with the graph?
2:38
Andy Powell -  want to see other players to be able to build stuff on it
2:39
Andy Powell -  encouaging people to think about "if you own metadata, what is it's place in the web of data?"
2:39
Andy Powelltalis.com/xiphos for full details
2:40
Andy Powell -  Q: what is your business model?   none currently - this is a prototype - but could think about pay-per-view - open access
2:40
Andy Powell -  PaulM: this work is about showing what is possible - new business models may emerge
2:41
Andy Powell -  Talis looking for/hoping for new data set that can be made more openly available
2:41
Andy Powell -  source code is available
2:41
Andy Powell -  or can be made available
2:42
Andy Powell -  Q: platform is a place to put data - but what stops it from becoming yet another silo - what still needs to be put in place to be able to work across different 'platform'-like services
2:43
Andy Powell -  robots will gather it all in - by following links in the data - open linked open data are working on this
2:44
Andy Powell -  Q: do tools exist right now?
2:44
Andy Powell -  ask Tom Heath :-)
2:44
Andy Powell -  Both Google and Yahoo are working in this space
2:45
Andy Powell -  Q: how are modifications to the graph treated - in terms of versioning?
2:45
Andy Powell -  stored in such a way that changes can be rolled back
2:47
Andy Powell -  next up...
2:47
Andy Powell -  Ian Corns, Talis - Project Zephyr: Letting Students Weave Their Own Path
2:48
Andy Powell -  disconnection of teaching/learning styles - students of google generation
2:49
Andy Powell -  digital natives - desire multimedia environment - always connected - real and virtual in parrallel
2:49
Andy Powell -  multi-plexing
2:50
Andy Powell -  disliked activities are simply skipped - huh??
2:50
Andy Powell -  they are content producers - particularly in a "rip, mix and burn" context
2:51
Andy Powell -  52% of first year undergrads are mature students (>21) but characteristics of google generation now more widely shared
2:51
Andy Powell -  now talking about resource lists - existing product is talis list
2:52
Andy Powell -  take a reading list from a lecturer and represent it electronically
2:52
Andy Powell -  access to resources more seemless for students - management of resources better for library
2:53
Andy Powell -  but - talis list not meeting the needs of google gen. students
2:53
Andy Powell -  project zephyr intended to overcome this
2:54
Andy Powell -  web of scholarly data - the reading list provides a significant way of linking resources - very interesting and valuable semantics
2:54
Andy Powell -  talis list is seen of a library application - major hurdles in getting lecturers to engage with it
2:55
Andy Powell -  student gets most benefit - costs lie with lecturer
2:55
Andy Powell -  need to find a way to provide lecturer with some value
2:55
Andy Powell -  zephyr intedned to do this
2:57
Andy Powell -  1st benefit to lecturers: by creating list in one place (zephyr) it can be surfaced in multiple places (facebook, library, etc.)
2:58
Andy Powell -  2nd benefit to lecturers: improve quality and depth of reading lists - e.g. see which resources have already been used elsewhere in other lists
2:59
Andy Powell -  3rd benefit: connecting the student with the lecturer - social networking - allowing student to ask questions, provide feedback, etc. - also connecting to peers
2:59
Andy Powell -  typical list has 500 items in it??   did he really say that!?
3:00
Andy Powell -  student experience... talis list is very much of the 'web of documents' vein
3:01
Andy Powell -  now showing screen shots of zephyr
3:02
Andy Powell -  lists presented thru navigable interface based on uni organisation hierarchy
3:02
Andy Powell -  having selected a list - student is shown a list of books and other stuff
3:03
Andy Powell -  can then filter by type
3:03
Andy Powell -  can also view superset of multiple list
3:04
Andy Powell -  can spot trends in terms of resources that are used in multiple modules
3:04
Andy Powell -  student can prioritise reading based on this
3:04
Andy Powell -  preserve structure of list as created by the lecturer
3:05
Andy Powell -  drill down into individual resources - including reviewing and rating by end-users
3:05
[Comment From Owen Stephens]
500 items not atypical - especially in humanities/arts (although sounds high as an average - science lists tend to be v short)
3:06
Andy Powell -  can also add resources into groups for collaborative activities, assignments, etc.
3:09
Andy Powell -  now looks like we have Nadeem Shabir (Talis) - not on programme i think
3:12
Andy Powell -  Note that Owen Stephens has been blogging day's talks at http://www.meanboyfriend.com/overdue_ideas/
3:12
Andy Powell -  presentation title is Open World Thinking
3:13
Andy Powell -  asking "what is the most widely used resource in first year undergrad comp sci courses?"
3:13
Andy Powell -  how can we answer that question?
3:13
Andy Powell -  data buried inside institutions
3:13
Andy Powell -  HE is silos within silos
3:15
Andy Powell -  institutions have been walled gardens - therefore solutions sold into institutions by third parties reflect that
3:15
Andy Powell -  the value in being open not recognised
3:15
Andy Powell -  not a technical problem
3:16
Andy Powell -  need a fundamental shift in thinking abouit openness
3:16
Andy Powell -  linked data is not just the semantic web done right - it is the web done right
3:17
Andy Powell -  designing for appropriation
3:17
Andy Powell -  us census data was made open without a specific application in mind
3:18
Andy Powell -  openness of description = coming to an agreed view on ways of describing things
3:18
Andy Powell -  then you can share stuff - integrate stuff - relate stuff
3:19
Andy Powell -  open descriptions + dereferenceable uris gives you interoperability for free
3:19
Andy Powell -  ontologies - formal representation of set of concepts within a domain
3:19
Andy Powell -  foaf - sioc - skos
3:20
Andy Powell -  e.g. the hybrid learning model (HML) (shown earlier) provides us with a simple ontology
3:21
Andy Powell -  Talis ontologies include...
3:22
Andy Powell -  Academic Institutions Internal Structures Ontology
3:22
Andy Powell -  Generic Lifecycle (workflow) Ontology
3:22
Andy Powell -  Resource List Ontology (sioc, bibo, foaf)
3:22
Andy Powell -  see www.vocab.org
3:22
Andy Powell -  Openness of access...
3:23
Andy Powell -  anywhere, anytime, anyhow (i.e. not necessary within a web browser)
3:23
Andy Powell -  this is a key to personalised learning
3:25
Andy Powell -  don't sell applications anymore - build contextualised views on web of data
3:26
Andy Powell -  it's only by being more open that the first question can be answered
3:26
Andy Powell -  openness is key to being able to rip, mix and burn
3:27
Andy Powell -  Xiphos is being built from ground up to embrace these kinds of questions
3:31
Andy Powell -  last presentation has finished - we are going into open discussion session - am going to sign off shortly

TrackBack

TrackBack URL for this entry:
http://www.typepad.com/services/trackback/6a00d8345203ba69e200e5533c75f18833

Listed below are links to weblogs that reference Talis Xiphos Research Day:

» Andy Powell - Web 2.0 and Repositories from Overdue Ideas
Andy has been live blogging at efoundations but clearly won't be doing this for his own session. [Read More]

Comments

Thanks for covering the day like this.

It's no small task to keep track of what is happening while summing up what you have just heard in to a nice distinct sentence, but you make it appear easy!

Thanks
Chris

Hi Andy - great live report on the day. Many thanks for this.

I'd very much agree with the remarks: "HE is silos within silos. institutions have been walled gardens - therefore solutions sold into institutions by third parties reflect that. the value in being open not recognised. not a technical problem. need a fundamental shift in thinking about openness."

And this is something we're attempting to address at this year's Institutional Web Management Workshop (IWMW 2008). This year's Innovation Competition (see http://www.ukoln.ac.uk/web-focus/events/workshops/webmaster-2008/competition/) will include data and APIs which is being made available by the Universities of Aberdeen, Bath and Edge Hill University.

The competition seeks to address some of the concerns you've raised: the data is often in walled silos and institutions don't recognise the value of being open. And addressing this is primarily a policy issues rather than a technical.

We're hoping that examples of lightweight development based on the policy decisions made at these three institutions will help to bring the "fundamental shift in thinking about openness".

If anyone is interested in attending the 3 day event in Aberdeen on 22-24 July, note that bookings close on Friday.
See http://www.ukoln.ac.uk/web-focus/events/workshops/webmaster-2008/
And further info on the Innovation Competition is available at
http://ukwebfocus.wordpress.com/2008/06/03/inn
ovation-competition-at-iwmw-2008/

Hi Andy, good to meet you yesterday and really enjoyed the talk. Trackback seems to be lost in the mist somewhere but my thoughts on your talk are at http://tinyurl.com/65y7o9.

Hi Andy, good to meet you yesterday and really enjoyed the talk. Trackback seems to be lost in the mist somewhere but my thoughts on your talk are at http://tinyurl.com/65y7o9.

Note that Owen Stephens also live-blogged the event in several posts to his blog starting here:

http://www.meanboyfriend.com/overdue_ideas/2008/06/talis-research-day---codename-xiphos.html

The comments to this entry are closed.

About

Search

Loading
eFoundations is powered by TypePad