W3C home > Mailing lists > Public > public-semweb-ui@w3.org > May 2007

Re: 90 minutes with HCI researchers - what would you discuss?

From: Hyowon Lee <hlee@computing.dcu.ie>
Date: Thu, 24 May 2007 00:08:37 +0100
Message-ID: <020c01c79d8f$4e9329e0$9a73ce88@Hyowonlaptop>
To: <Lloyd.Rutledge@cwi.nl>, <public-semweb-ui@w3.org>

>  "In general, HCI methodology has three phases: analyzing your users'
>  needs, designing an interactive system that meets those needs and
>  evaluating the system you build using this design. While this approach
>  typically applies to building specific systems, its role in research
>  raises certain issues. One is that research tends, rightly or wrongly
>  (or perhaps due to the nature of research), to be more technology
>  driven, with researchers wishing to test hypotheses about interactive
>  possibilities that specific new technologies enable. Here, the
>  challenge is to make sure any solutions explored have a legitimate
>  user problem to solve, that is does indeed solve it, and that it
>  solves it better than other available solutions for the same
>  problem. Thus, even with a particular solution technology in mind, the
>  Semantic Web researcher still needs to apply the main three HCI
>  phases: finding a user base for their technology and analysing their
>  requirements, designing a system that uses this technology to meet
>  these requirements and then evaluating how well the technology
>  addresses these requirements in real use. In each phase, the
>  technology that the research is applying should be compared with other
>  available technologies that can address the same problem."
> 
> People who are more familiar with HCI than me, please make any
> appropriate corrections or additions.
> 
> I hope this helps frame discussion of how to perform HCI in SW
> research.  Perhaps the "technology driven" discussion is
> controversial.


I think the text is generally true and useful for us.  But I just want to raise one thing about:

> Here, the challenge is to make sure any solutions explored have 
> a legitimate user problem to solve, that is does indeed solve it, 
> and that it solves it better than other available solutions for the same
>  problem. 

Maybe the above should not be too emphasised?  I mentioned in this mailing list before about how HCI problem of user annotation burden for personal photo management can be nicely reduced by having automatic face detection technique - surely in this case (and probably many other cases) it follows:

 - Problem: too much time & effort to manually annotate the faces in my (thousands of) photos
 - Solution: use automatic face annotation tool

On the other hands, we also simply come up with something new (very technology-centric way) - not knowing whether or where this could be useful, or what problem this would solve - then play with it, show it to people, maybe use it for some time, and come up with *new use* for it.  New technology is not just solving the existing problems, it can do something we didn't even consider as problems.

Sometimes I have to just show our new interface features to people and see what they think - users don't usually come up with new ideas in use, they are good at identifying usability problems.  Often it doesn't work if we visit a user and ask "is there any new feature that you want us to provide?"

For example, keyframe-based (spatial) video browsing is currently a norm in most video retrieval systems, but when it first appeared, it didn't come out as a result of careful analysis of user needs or requirements - it was born of technical approach called content-based image analysis, which automatically segments video sequence into individual camera shots, then extracting one frame from each shot.  Once this is done, it was possible to present a number of thumbnail-size keyframes on a single screen, and it proved to be very useful - not having to play or FF/ REW, but just see the whole thing at one glance, so effectively, get the rough content view of a video.

I guess maybe we can say that the users had problem of having to see the video content quickly but couldn't (because video is time-based medium), but I don't think any user would have even questioned "is there some faster and easier way to get an overview of a video without playing it?"

Surely the detailed design issues of exactly how thumbnail keyframes should be presented, in what size, how many per screen, etc. that the HCI practice should find out, but my point is that there was no clearly defined problems that invented this very useful feature.

Hyowon

----------------------------------------------------
Dr Hyowon Lee
Post-doctoral researcher
Centre for Digital Video Processing
Dublin City University
Glasnevin, Dublin 9, Ireland
Tel: +353 -1 -7005829
http://www.computing.dcu.ie/~hlee/
Email: hlee@computing.dcu.ie
----------------------------------------------------


----- Original Message ----- 
From: "Lloyd Rutledge" <Lloyd.Rutledge@cwi.nl>
To: <public-semweb-ui@w3.org>; "Hyowon Lee" <hlee@computing.dcu.ie>
Sent: Wednesday, May 23, 2007 1:23 PM
Subject: Re: 90 minutes with HCI researchers - what would you discuss?


> 
> For some discussion of some SW techniques relating to search, see the
> survey that Michiel Hildebrand has started in an earlier posting on this
> list.
> 
> In addition, I added to
> http://www.webscience.org/swuiwiki/index.php?title=How_To#Methodology
> some discussion of what HCI techniques mean in the context of Semantic
> Web research.  The new text is:
> 
>  "In general, HCI methodology has three phases: analyzing your users'
>  needs, designing an interactive system that meets those needs and
>  evaluating the system you build using this design. While this approach
>  typically applies to building specific systems, its role in research
>  raises certain issues. One is that research tends, rightly or wrongly
>  (or perhaps due to the nature of research), to be more technology
>  driven, with researchers wishing to test hypotheses about interactive
>  possibilities that specific new technologies enable. Here, the
>  challenge is to make sure any solutions explored have a legitimate
>  user problem to solve, that is does indeed solve it, and that it
>  solves it better than other available solutions for the same
>  problem. Thus, even with a particular solution technology in mind, the
>  Semantic Web researcher still needs to apply the main three HCI
>  phases: finding a user base for their technology and analysing their
>  requirements, designing a system that uses this technology to meet
>  these requirements and then evaluating how well the technology
>  addresses these requirements in real use. In each phase, the
>  technology that the research is applying should be compared with other
>  available technologies that can address the same problem."
> 
> People who are more familiar with HCI than me, please make any
> appropriate corrections or additions.
> 
> I hope this helps frame discussion of how to perform HCI in SW
> research.  Perhaps the "technology driven" discussion is
> controversial.
> 
> -Lloyd
> 
> Hyowon Lee wrote:
>> It will be useful at the beginning of the SWUI workshop if somebody
>> showed us a list of key SW techniques achieved so far or being
>> developed now - for a person like myself who doesn't know SW in
>> depth/breadth, and for us to get a summary of SW development.
>>
>> It will help us to think about and maybe map the kinds of UI & HCI
>> elements suitable for the underlying SW techniques.
>>
>> Hyowon
>>
>> ----------------------------------------------------
>> Dr Hyowon Lee
>> Post-doctoral researcher
>> Centre for Digital Video Processing
>> Dublin City University
>> Glasnevin, Dublin 9, Ireland
>> Tel: +353 -1 -7005829
>> http://www.computing.dcu.ie/~hlee/
>> Email: hlee@computing.dcu.ie
>> ----------------------------------------------------
Received on Wednesday, 23 May 2007 23:06:20 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:16:46 GMT