Social media, Ontologies and Web 2.0 eLearning Paola Monachesi (Utrecht University) work carried out in collaboration with Kiril Simov, Petya Osenova, Eelco Mossel, Vlad Posea, Thomas Markus
May 11, 2015
Social media, Ontologies and
Web 2.0 eLearning
Paola Monachesi (Utrecht University)
work carried out in collaboration with Kiril Simov, Petya Osenova, Eelco Mossel,
Vlad Posea, Thomas Markus
Overview
• Ontologies for eLearning– Lexicalized ontologies for cross-lingual retrieval of
learning material– Ontologies vs. Tagging/Folksonomies– Integrating ontologies and tagging for knowledge
discovery– Integrating ontologies and social networks for
knowledge discovery
• Evaluations and challenges• Conclusions
Ontologies
• Ontologies are a crucial element of the Semantic Web vision
• Ontologies allow for a formalization of knowledge that:– facilitates automatic processing of the
information;– enables inference to be performed.
Ontologies and eLearningExamples of two possible uses:
• enhance the management, distribution and retrieval of the learning material – LT4eL project (www.lt4el.eu)
• ontologies enriched with social tags can mediate between formal and informal learning– LTfLL project (www.ltfll-project.org)
Ontologies and eLearning
• Users:– Tutors/content providers who want to compile
a course– Learners that want to find material in several
languages – Learners that are looking for content in a
knowledge discovery process and for peers
Demo
Enhancing the retrieval of multilingual learning material:
http://www.lt4el.eu/index.php?content=videos
Components
• A corpus of learning objects in 8 languages
• A domain ontology
• Lexicons for 8 languages
• (Linguistically, semantically) annotated learning objects
LT4eL Domain Ontology: general issues
• The domain: Computing• Coverage: operating systems; programs; document
preparation – creation, formatting, saving, printing; Web, Internet, computer networks; HTML, websites, HTML documents; email
• The role of the ontology: for indexing of the LOs
Connection with other Ontologies
DOLCE (Guarino&a
l.)
OntoWordNet
LT4EL
Current state of the ontology
• about 1002 domain concepts,
• about 105 concepts from DOLCE
• about 169 intermediate concepts from OntoWordNet
• http://www.lt4el.eu/index.php?content=tools#ontology
Ontology-Based Lexicon Model
• The lexicons represent the main interface between the user's query and the ontology
• Lexicons for all languages (8) of the project have been created
Mapping Lexical Varieties
Ontology
LexicalizedTerms
Free Phrases
<entry id="id60"> <owl:Class rdf:about="http://www.lt4el.eu/CSnCS#BarWithButtons"> <rdfs:subClassOf> <owl:Class rdf:about="http://www.lt4el.eu/CSnCS#Window"/> </rdfs:subClassOf> </owl:Class> <def>A horizontal or vertical bar as a part of a window, that contains buttons, icons.</def> <termg lang="nl"> <term shead="1">werkbalk</term> <term>balk</term> <term type="nonlex">balk met knoppen</term> <term>menubalk</term> </termg></entry>
Lexicon Entry
Ontology and Multilingual Data
EN
DE
DT
Lexicons Documen
ts
Ontology
DT
DE
EN
Annotation of LOs
• Annotation of the text with concepts– Identification of the text chunk that will be
annotated
– Assigning of all possible concepts for the chunk
– Concept disambiguation
1. Better retrieval of LOs– Find LOs that would not be found by simple text search (where
exact search word must occur in text)
2. Multilinguality– One implementation applies to all languages in the project
3. Crosslinguality– Possible to find LOs in languages different from search/interface
language• No need to translate search query
• Search possible with passive foreign language knowledge
Added Value
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Full text search
Keyword search
Semantic search
Concept browser
Definition finder
Success in answering quiz questions by functionality
Functionality
Success
Student: Searching Test 2a Results
(target groups only)
Student: Searching Opinions 2
• What did they dislike about Semantic Search?
– It didn't return relevant results.– because it doesn't find what I am searching for– its too vague– i didn't use it a lot as the results were chaotic– i find it not much to the point for the types of research i
usually do– It is a bit too much to offer this much search methods– the name semantic is confusing– i liked this type best.. it was the easier to find the relevant
information
Student: Searching Comments: Semantic
• What did they dislike about the Concept Browser?
– It didn't return relevant results– It was too slow for my part and did not give any additional
value– it's a roundabout way of searching– was not eay to use it. maybe this was because i did not
fully understand how the concept browser worked– I am not sure what the concept browser is– don't know what it is.– for content questions this might be a relevant search
method. However, less relevant when studying a language– I like that method the most but it wouldn't be useful in my
studies - English philology– it helps a lot to understand given topic/term
Student: Searching Comments: Concept Br
0% 10% 20% 30% 40% 50%
full text search
keyword search
semantic search
concept browser
definition finder
Most useful functionality - students' opinions
Functionality
Percentage finding this the most useful
Student: Searching Opinions 1
Ontologies and social media
• Ontologies– can support the learner in the learning path; – provide the formalization of domain knowledge
approved by expert (Monachesi et al. 2008)
• Challenges– Too static;– Incomplete;– Knowledge acquisition bottleneck– Mismatch with the view of a domain by a learner – Tagging might provide better representation
Ontologies and social media
• Aim:– Create a link between the formal
representation of a given domain in the form of ontologies
and– The informal description produced by social
tagging and folksonomies
Tagging
• Main issues:– Add the informal dimension to learning by including
learning material from social media and tags• Videos (Youtube)• Images (Flickr) URLs to relevant websites (Delicious)• Q&A (Yahoo answers) Forums Blogs
– Employ NLP techniques to extract domain knowledge and relations from learning material and tags
– Create a link between existing domain ontologies, social tagging and learning material
Ontology enrichment with social tagging
• Exploit tagging to access and extract knowledge from social media applications
• Establish a link between tags, concepts and resources.
• Investigate impact of enriched ontology on advanced learners and beginners
Ontology enrichment with tagging
Experiment with delicious.com data
Social media application: Delicious• Assess:
– Whether it is possible to find related tags in case of limited resources and users, as in the case of eLearning application.
• Use of tag co-occurrence• (Use of cosine similarity)
– How the related tags corresponds with concepts present in the ontology
Tri-partite model
Tags Users Resources
Experiment with delicious.com data
• Use a domain ontology on computing• From most popular tags of delicious, select those that are in the
ontology• Top 5 tags found in this way:
– design– blog– tools tool– software– linux
• Experiment for 1 tag at a time• Select bookmarks for the tag• Find related tags by co-occurrence
Criteria for selection of bookmarks
• 5 classes of numbers of users who tagged the resource:– A: 8-13– B: 14-25– C: 26-50– D: 51-100– E: 101-200
• For each class selected 15 most recent bookmarks, for which holds:– Seed tag occurs in top-5 tags for the bookmark– Saved by the desired number of users
Example data for a bookmark
bookmarked url = wiki.mindtouch.com/MindTouch_Deki
number of people who saved this bookmark = 20
Tag: Number of users that assigned the tag:
Windows 16
Screenshot 10
Freeware 7
Utility 6
Software 6
Tool 5
Free 5
Image 3
Application 3
Work 1
Results: Related tagssoftware software:
delicious top-11
(gold standard)
linux linux:
delicious top-11
(gold standard)
Class A: 8-13 users
(sample of 15 bookmarks)
windows tools
windows
opensource
programming
mac
web
free
freeware
web2.0
utilities
linux
howto
Ubuntu
ubuntu
howto
tutorial
software
sysadmin
unix
opensource
reference
security
tools
programming
Class B: 14-25 users
(sample of 15 bookmarks)
free
freeware
windows
howto
reference
ubuntu
Class C: 26-50 users
(sample of 15 bookmarks)
freeware
tools
howto
reference
software
ubuntu
Class D: 51-100 users
(sample of 15 bookmarks)
free
freeware
mac
*macosx*
*mobile*
*osx*
howto
opensource
software
sysadmin
tutorial
ubuntu
Class E: 101-200 users
(sample of 15 bookmarks)
free
freeware
mac
tools
windows
howto
opensource
software
tutorial
ubuntu
delicious tags vs. computing ontology
Related tags for
softwareRelated tags for
linuxMerged: all related tags
for top-5 selected tags:
design, blogs, tools, software, linux
freeware mac tools (in ontology: tool) windows
macosx
mobile
osx
free (no CS)
= in ontology
software ubuntu
howto
opensource
reference
sysadmin
tutorial
= in ontology
• 33 related tags found• 7 of 33 are not in domain• 26 of 33 are in domain (79%)• 23 of 26 are in gold standard (88%)• 13 of 26 are in ontology (50%)
Aspects of ontology enrichment
• Mapping of related tags to existing concepts– Tag as concept– Tag as lexicalization
• Manual process but working towards heuristics for automatic assignment
• Addition of relations
Tag relation to concept relationFound relations between tag software, and other tags that are in the ontology:
Ontology integration for knowledge discovery
User evaluation
• Assumption: Enriched ontology can be a valid support for knowledge discovery given the explicit relations between concepts vs. tag visualization
• Hypothesis: differences in knowledge discovery approach (advanced vs. beginners)– Beginners: prefer tag visualization– Advanced: prefer ontology
Setup
• Learning task: quiz solving on markup languages – 3 questions to be answered with ontology enhanced
with tags– 3 questions to be answered with tags
• 6 beginners (no CS background, no knowledge of the domain)
• 6 advanced (CS background)• Results: questionnaire with 10 questions
Results - beginners
• What was useful in finding the answer by using:– the enriched ontology
• documents (4.4)• social tags (3.6)• conceptual structure (2.8)
– the clouds of tags• documents (3.4)• related tags (2.8)• structure (1.8)
Results -advanced learners
• What was useful in finding the answer by using:– the enriched ontology
• social tags (4.33)• conceptual structure (3.17)• documents (3.17)
– the clouds of tags• documents (3.5)• structure (3.0)• related tags (3.0)
Summary
• Beginners prefer documents rather than structure
• Advanced learners rely on tags and structure more than beginners
Social networks
• Main issues:– Knowledge discovery based on social
networking– Adapt search and recommendation algorithms
for finding relevant peers and resources– Support networking for learning purposes
Research issue
• Communities of users with common interests use multiple social networking applications
• Can we offer support?– Support = personalized search and
recommendations across social networking applications
Architecture of the applicationData sources
Indexing mechanism for the data produced in the user network
repository
Support through personalized search and recommendation
User interface
Design of services• Adapted FolkRank algorithm for search and
recommendation based on tags• Search by disambiguating tags using
knowledge bases (DBpedia and Freebase)• Convert information extracted from the social
network into semantic friendly formats (FOAF, SIOC, SCOT, DC)
Future work • Social network based knowledge discovery
– Social networks and tags
Integrated with• Content based knowledge discovery
– Ontology enhanced with tags
Integrated with • Formal learning
– Semantic annotated documents with discourse
CSF
Common Semantic Framework
• Objectives– support formal and informal learning and the
emergence of new knowledge– communication among users– identification, retrieval and recommendation
of relevant material
Support formal and social learning: Goal
• To support:– Knowledge discovery– recommendation of formal and informal
learning material and users• By means of:
– ontologies – tagging– social networks
Architecture
Conclusions
• Ontologies enriched with tags have a potential to support knowledge discovery
• Challenges:– Visualization– Best way to integrate the two– Role of social networks