SPEECH COMMUNICATION SCIENCES &
ADVANCED COMPUTING IN THE HUMANITIES
Working Group III on Educational Materials and Tools
Report from the Prague meeting, March 17-19, 2000
Coordinator: Koenraad de Smedt
Subgroup coordinators:
-
BILL BLACK (CL)
-
EVA WIGFORSS (CD)
-
MIKE MCTEAR (SLE)
-
VALERIE HAZAN (PHON)
-
JACQUES SOUILLOT (NEL)
Working group members:
Concerns
Providing information on various kinds of materials for education and training
available in electronic form. This information will be provided by means
of a searchable and extensible database.
Material types
The following types of materials are considered for inclusion in the database:
-
Course or tutorial (self-contained, electronic on the Internet or CD-ROM)
-
Course notes (not part of course)
-
Exercise set (not part of course)
-
Demo
-
Tool (usable for training, also kit of several tools, also hardware)
-
Corpus
-
Lexicon
-
Grammar
-
Textbook (printed)
-
E-journal (with educational relevance)
-
Inventory (bibliography, index, e-library, etc.)
Material descriptions
These materials will be listed in a database that is browsable and searchable
on the web. Suggestion for database fields for each item:
-
Name of the item (free text)
-
Medium (menu: CD-ROM, Printed or Web)
-
Where available? (publisher and/or URL)
-
Author (author name + institution + email; multiple values when multiple
authors)
-
Type (menu of the material types listed above)
-
Platform requirements (free text)
-
Subject field (menu: SLE, Phonetics, CD, CL, NEL)
-
Content (choose one or more keywords: automata, parsers, grammar, NLP applications,
information retrieval,... or enter new keyword)
-
Language of presentation (menu)
-
Languages that are the object of study (menu, multiple values)
-
Main features with respect to pedagogy, interface and/or navigation (choose
one or more keywords: multiple-choice exercises, open exercises, solutions
to exercises, illustrations, student tracking, notes, index, references,
... or enter new keyword)
-
Audience (free text, e.g. first-year students of computational linguistics)
-
Description (free text, e.g. many realistic exercises)
-
Shortcomings and caveats (free text, e.g. not tested on significant student
numbers)
-
Illustration (URL to screen shot or other illustration)
-
Availability (menu: free, free registration, or paid)
-
Associated materials (reference to other database entries; multiple values)
-
Date of entry (some standard date format)
Not all fields are obligatory, e.g. illustration is optional. Unless
a field has free text, consensus must be reached on a set of keywords (e.g.
avoid that one entry has "multiple choice ex." while the other has "MC
exercises"), although the list of keywords must be dynamically extensible.
Translations of the keywords and fields to different languages will be
considered. For every field, there will be help and suggestions or examples.
There must be room for feedback on desired new fields.
Data collection
Proposal for initial data collection: the initial database is filled with
descriptions of some relevant materials collected by the subject expert
groups for the different fields. By way of testing, at least some materials
will be coded separately by different people and the resulting entries
will be compared.
Proposal for self-sustainability: authors (i.e. anybody, in practice)
must be able to register their materials themselves. Entry by authors is
preferable to entry by others. There will be no limitation to Europe.
Proposal for comments: users must be able to enter user comments. Comments
must be signed. There will be no rating system. There will be no requirement
to log in or register before accessing the database or giving comments.
Searchability
The database must be searchable/browsable in different ways.
-
Basic search/browsing is by choosing one or more types (from a menu), one
or more content keywords (from a keyword list, either alphabetical or hierarchical),
and/or one or more subject fields (from a menu). The resulting list of
results can be browsed or further restricted.
-
Advanced search is by placing restrictions on any number of fields.
This search mask will be basically similar to the data entry mask.
-
Free text search is by searching all fields with keywords (easy but not
very reliable)
Inventory of computational resources
-
Computational linguistics and human language technologies
-
Consider also computational tools and resources in other linguistic studies
(general linguistics, corpus linguistics, psycholinguistics, languages)
-
Speech Communication Sciences (CAL/Internet Resource Inventory: SCS Book
1, Chapter 5, Appendix 1; Inventory database on Web: http://www.phon.ucl.ac.uk/resource/cal-top.htm)
-
Phonetics (Survey of textbooks and equipment: SCS Book 2, chapter
2.3)
-
Spoken Language Engineering (Textbook and other suggestions in SCS Book
2, chapter 3.3; Web tutorials and other materials: SCS Book
3, chapter 3; Converted by the DISC project to Web at: http://www.disc2.dk/tools/Relevantweb.html
)
-
Speech and Language Therapy
-
See also demos at the Linglink site http://www.linglink.lu/hlt/showcase/
Further work
We must establish a web site and an email list for coordinating further
work. This will mainly be the responsibility of the WG coordinator.
A prototype data entry mask will be made, possibly based on an XML skeleton.
The subgroup coordinators will be responsible for making sure that example
codings are established for some materials within their field of expertise.
They will further divide work within their subgroups and may in some later
stage also contact the authors of materials in order to stimulate them
to do data entry themselves.
P.S. The following task has been transferred to the supervision of WG
I.
General pedagogical and technical considerations
There will be a synthesis of earlier formulated pedagogical considerations
for building and using these resources (CAL group, SCS Book 1, chapter
5; Book 2, chapter 6, Book 3, chapter 5;
ACO*HUM
Book, chapters 4 and 5; and other material).