15
Crosslingual Retrieval in an eLearning Environment Cristina Vertan, Kiril Simov, Petya Osenova, Lothar Lemnitzer, Alex Killing, Diane Evans, Paola Monachesi http://www.lt4el.eu/ Artificial Intelligence and Human- Oriented Computing (September 10-13, 2007 - Roma, Italy)

Crosslingual Retrieval in an eLearning Environment

Embed Size (px)

DESCRIPTION

Crosslingual Retrieval in an eLearning Environment. Cristina Vertan, Kiril Simov, Petya Osenova, Lothar Lemnitzer, Alex Killing, Diane Evans, Paola Monachesi http://www.lt4el.eu/ Artificial Intelligence and Human-Oriented Computing (September 10-13, 2007 - Roma, Italy). Framework. - PowerPoint PPT Presentation

Citation preview

Page 1: Crosslingual Retrieval in an eLearning Environment

Crosslingual Retrieval in an eLearning Environment

Cristina Vertan, Kiril Simov, Petya Osenova, Lothar Lemnitzer, Alex Killing, Diane Evans,

Paola Monachesihttp://www.lt4el.eu/

Artificial Intelligence and Human-Oriented Computing

(September 10-13, 2007 - Roma, Italy)

Page 2: Crosslingual Retrieval in an eLearning Environment

Framework

• EU-Project LT4eL: Language Technology for eLearning (www.lt4el.eu) 12 Partners

• Learning objects: in 9 Languages• eLearning Test-System : open source

platform ILIAS (www.ilias.de/)• Domain: Computer Science for non CS

specialists• Services: Keyword Extraction,

Definition Detection, Semantic Search

Page 3: Crosslingual Retrieval in an eLearning Environment

Architecture (2)

Place of multilinguality - real situation

Lexikon

CZ

CZCZEN

ENCONVERTOR 1

Documents SCORM

Pseudo-Struct.

Basic XML LING. PROCESSOR

Lemmatizer, POS, Partial Parser

CROSSLINGUAL RETRIEVAL

LMSUser Profile

Documents SCORM

Pseudo-Struct

Metadata (Keywords)

Ling. Annot XML

Ontology

CONVERTOR 2

Documents HTML

Lexikon

PT

Lexikon

RO

Lexikon

PL

Lexicon

GE

Lexikon

MT

Lexikon

BG

Lexikon

DT

Lexicon

EN

PLPL

GEGE

BGBG

PTPT

MTMT

DTDT

RORO

ENDocuments User

(PDF, DOC, HTML,

SCORM,XML)

REPOSITORY

Glossary

Page 4: Crosslingual Retrieval in an eLearning Environment

• Processing of the keywords in LOs• Formalization of the meanings –

definitions from Internet• Linking to an upper ontology

(DOLCE)• Addition of new concepts• Addition of relations• Documentation• Lexicons in 9 languages

The Creation of LT4eL Ontology

Page 5: Crosslingual Retrieval in an eLearning Environment

Connection with other Ontologies

DOLCE (Guarino&a

l.)

WordNet

LT4EL

Page 6: Crosslingual Retrieval in an eLearning Environment

<owl:Class rdf:about="http://www.lt4el.eu/CSnCS#WebPage">

<rdfs:comment>A document (file) connected to the

World Wide Web and viewable by anyone connected to

the internet who has a web browser.</rdfs:comment>

<rdfs:comment>Hyper CSnCS:

http://www.lt4el.eu/CSnCS#TextFile</rdfs:comment>

<rdfs:comment>Equal WN20: ENG20-05964213-n</rdfs:comment>

<rdfs:comment>ID: id1757</rdfs:comment>

<rdfs:subClassOf>

<owl:Class rdf:about="http://www.lt4el.eu/CSnCS#TextFile"/>

</rdfs:subClassOf>

</owl:Class>

Ontology Example

Page 7: Crosslingual Retrieval in an eLearning Environment

<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

Page 8: Crosslingual Retrieval in an eLearning Environment

Ontology and Multilingual Data

EN

DE

DT

Lexicons Documen

ts

Ontology

DT

DE

EN

Page 9: Crosslingual Retrieval in an eLearning Environment

• Starting points– A multilingual document collection

– An ontology including a domain ontology on the domain of the documents

– Concept lexicalisations in various languages

– Annotation of concepts in the documents

Multilingual, Semantic Document Retrieval

Page 10: Crosslingual Retrieval in an eLearning Environment

• Improved access to documents– Find docs that would not be found by

simple text search

• Multilinguality– One implementation for multiple

languages

• Crosslinguality– Retrieve documents in languages other

than language of the query or ontology presentation

Goals of the Approach

Page 11: Crosslingual Retrieval in an eLearning Environment

The User1. Submits a free text query2. Sees document list

A list of documents is displayed with some meta information, for example:

• title; length; original language; • keywords and concepts that are common to both the query

and the document; • other keywords and concepts that are related to the document

but not to the query

3. Sees concepts for refining searchConcepts related to the search query starting

point for browsingNo related concepts from search query root of

ontology starting point

Outline of Search Procedure

Page 12: Crosslingual Retrieval in an eLearning Environment

4. Views documentsUser looks into the documents from the list and

estimates their relevance

5. Browses ontologyEntry point depends on the initial query

6. Selects conceptsConcepts are used for query refinement

7. Selects search optionThe search option is about how to use the

ontology fragments for search

Outline of Search Procedure

Page 13: Crosslingual Retrieval in an eLearning Environment

8. Sees new document list A new list of documents is displayed, based only on

ontological search

9. Sees updated concept browsing unitsConcepts that are common to the found documents

Example: Concept “Report” Some documents about academic

writing Concept “Publication”

10.Repeats steps from step 5 (Browse ontology)

User selects another set of related concepts and submits it as the search key, etc

Outline of Search Procedure

Page 14: Crosslingual Retrieval in an eLearning Environment

Integration in ILIAS

Page 15: Crosslingual Retrieval in an eLearning Environment

We are ready to• Compare different searches over

LOs• Tune parameters of the searches• Validate the user added value of

the provides services• Add relations and extend the

search• Provide context of the result from

the search

Conclusion and Future Work