An Usability Study Of A Recommender System For Mash Ups For Learning

ReMashed –An Usability Study of a
Recommender System
for Mash-Ups for Learning
Hendrik Drachsler, Dries Pecceu, Tanja Arts, Edwin Hutten, Peter van Rosmalen,
Hans Hummel & Rob Koper

My background…

Learning Networks

hendrik.drachsler@ou.nl

International Conference on Interactive Computer Aided Learning, ICL 2009,Villach, AT
Page 2 | 23 September, 2009

Learning Networks
Explicitly address informal
learning

Learners can publish, share,
rate, tag and adjust their
own Learning Activities in a
Learning Network

Open Corpus that emerges
form the bottom upwards



Personal Environments
Nowadays …

More Information
Blog Reader Providers

Social
Bookmarking
Various
Communities



Personal Learning Environments



Selection problem because …

…of the amount
of data that is
emerging.

…learners can be
overwhelmed by
the plethora of
information.


Can we create a Recommender System
Today, Recommender Systems
for Mash-Ups for Learning?
supporting our decisions



What is ReMashed?

A Mash-up environment that allows you to
personalize emerging information of online
communities with a recommender system.

You tell what kind of Web 2.0 services you
use and then you are able to define which
contributions of other members you like
and do not like.


Goals for ReMashed
1. End-User level
Providing a recommender system for Web 2.0
sources of learners in informal Learning
Networks.
2. Researcher level
1. Offering researchers a system for the evaluation of
recommendation algorithms for learners in informal
Learning Networks.
2. Creating user-generated-content data sets for
recommender systems in informal Learning Networks.


How does it work?

ReMashed uses collaborative filtering to
generate recommendations.

It works by matching together users with
similar tastes (neighbours) on different
Web 2.0 resources (delicious, Flickr, blog
feeds, Slideshare, Twitter, and YouTube).



How does it work?

Cold-Start = Tag-based recommendation

Collaborative Filtering with ratings



The 1st Release



The 1st Release

Recommendation
Algorithms

Database User Interface
of Items



System Evaluation
• At the TENCompetence Winterschool 2009

• Also external users sign up for the evaluation phase

• In total 49 people from 8 different countries
After this slide you
• created some
The evaluation phase ran for one month and was concluded
with an online recall questionnaire. really ugly ones do
you really want to
show them to the
• We received answers from 19 participants in total
audience?
(response rate of 38%).



Usage of Web 2.0 services
Strongly Agree Neutral Disagree Strongly
agree disagree
16% 5% 31% 11% 37%
1. Use a blog 21%
(n=3) (n=1) (n=6) (n=2) 48% (n=7)

2. Use social 42% 42% 6% 5% 5%
Bookmarking 84%
(n=8) (n=8) (n=1) (n=1) 10% (n=1)

5% 32% 21% 26% 16%
3. Share pictures 37%
(n=1) (n=6) (n=4) (n=5) 42% (n=3)
They just
4. Share 11% 26% 21% 21% twittered
17%
presentations 37%
(n=2) (n=5) (n=4) (n=4) 38% (n=5)
that the
5. Use micro- table is
11% 21% 10% 21% 37%
blogging
32%
(n=2) (n=4) (n=2) (n=4) 58% terrible.
(n=7)

0% 15% 38% 26% 21%
6. Upload videos 15%
(n=0) (n=3) (n=7) (n=5) 47% (n=4)

7. Use YouTube 5% 32% 15% 16% 32%
to collect videos 37%
(n=1) (n=6) (n=3) (n=3) 48% (n=6)

Satisfaction with ReMashed
Very Satisfied Unsatisfied Very Total
satisfied unsatisfied
1. Overall 5% 58% 26% 11% 100%
satisfaction (n=1) 63% (n=11) (n=5) 37% (n=2) (n=15)
2. Satisfaction tag-
20% 40% 27% 13% 100%
based algorithm
(n=3)
60% (n=6) (n=4)
40% (n=2) (n=16)
in the beginning
3. Satisfaction tag-
0% 69% 19% 13% 100%
based algorithm
(n=0) 69% (n=11) (n=3) 31% (n=2) (n=16)
at the end
4. Satisfaction
rating-based 8% 53% 31% 8% 100%
algorithm in the (n=1) 61% (n=7) (n=4) 39% (n=1) (n=14)
beginning
5. Satisfaction
rating-based 8% 54% 31% 8% 100%
algorithm at the (n=1) 62% (n=7) (n=4) 38% (n=1) (n=13)
end

Open Question
We received 9 answers:

• Integrate social networks like Linkedin, Facebook,
and MySpace.

• Integrate mind mapping tools like Mindmeister.

• Create clusters of Web 2.0 services of the same type,
e.g. combine Picassa and flickr in a subcategory
‘Pictures’.



The 2nd Release



The 2nd Release

DUINE Prediction
Engine

Database User Interface
of Items



You can use it as well!
Register at ReMashed
remashed.ou.nl. starts mashing.

Enter your favorite Taste your
Web 2.0 potatoes. personal
flavor of
Web 2.0.
Join the
community.



Future R&D
End-user perspective:
1. More Web 2.0 services and social networks
2. Widget / Portlet interface for other Personal Environments
3. Administration interface to create instances of the
ReMashed

Researcher perspective:
1. Creating data sets for Technology-Enhanced Learning
2. Exploring new recommendation technologies
3. Creating web services to offer recommendations to other
Mash Up Personal Learning Environments



Many thanks for your interest!

This slide is available here:
http://www.slideshare.com/Drachsler

Email: hendrik.drachsler@ou.nl

Skype: celstec-hendrik.drachsler

Blogging at: http://elgg.ou.nl/hdr/weblog

Twittering at: http://twitter.com/HDrachsler



Please sign up at:

http://remashed.ou.nl



An Usability Study Of A Recommender System For Mash Ups For Learning

More Related Content

Similar to An Usability Study Of A Recommender System For Mash Ups For Learning (20)

More from MegaVjohnson (15)

Recently uploaded (20)

An Usability Study Of A Recommender System For Mash Ups For Learning