TECHNIQUES FOR KNOWLEDGE SHARING IN THESIS REVIEWING

(1)

Prof. Dr. Wolf-Fritz Riekert

Hochschule der Medien (HdM) Stuttgart University of Applied Sciences

mailto:riekert@hdm-stuttgart.de http://v.hdm-stuttgart.de/~riekert

TECHNIQUES FOR KNOWLEDGE SHARING IN THESIS REVIEWING

I-KNOW ‘05

Graz, Austria, June 29 – July 1, 2005

TECHNIQUES FOR KNOWLEDGE SHARING IN THESIS REVIEWING 2

THESIS REVIEWS

z Reviewing theses is a knowledge task

z Knowledge in thesis reviews not adequately used Öretrospective: of less interest for the thesis author Öconfidential: not to be disclosed to other people

z Main idea: New ways of sharing the knowledge through a computer-based system

ÖSupport the task of the reviewer

ÖGenerate personalized thesis reviews for the authors ÖDerive generalized teaching material for future authors

FIRST APPROACH:

AN ELECTRONIC FORM Acrobat PDF form supports reviewing and grading of bachelor and master theses

z 20 success criteria can be ticked as positive, negative or neutral

z Points can be given for 4 categories of success criteria:

ÖContent (max. 50 pts) ÖComposition (max. 20 pts) ÖForm (max. 15 pts)

ÖCitation style (max. 15 pts)

z Gradation automatically computed (1 = very good, ... , 5 = failure)

z Review summary

TOWARDS A DATABASE DRIVEN SOLUTION

z Shortcomings of the simple electronic form approach:

ÖThesis reviews are unintelligent isolated documents ÖOnly one large comment can be formulated: the review

summary

ÖNo possibility to formulate multiple specific comments ÖGranularity of represented knowledge too coarse

z Therefore: database driven approach

z Entity-relationship modeling of reviewing knowledge

z Separate entity types for Ötheses,

Öcomments, Öcriteria and

Öcategories of criteria

(2)

ENTITY-RELATIONSHIP MODELING OF REVIEWING KNOWLEDGE

thesis

fulfillment indicator

text

location comment

category

criterion

one of “content”, “composition”,

“form”, and “citation style”

e.g.: “figures complete”,

“all technical terms defined”

e.g.: “−” (negative) e.g.: “legend missing in fig. 2.3”

e.g.: “page 27”

1 n

IMPLEMENTATION

The system is implemented as a MS Access database application

z Form-based data entry

Ömanagement of multiple theses

Öunlimited collection of comments per thesis Öpull-down menu for criteria

z Automated report generation

ÖIndividual review report (for the author)

Ö„Slide Show“ contains general knowledge (for other reviewers and students)

DATA ENTRY FORM

REPORTS

z Individual review report (for the author)

z „Slide Show“ contains general knowledge (for other reviewers and students)

(3)

SLIDE SHOW

z merges comments from a number of theses

z grouped by category, each category starts a new slide

z ordered by success criteria within category Öelicits typical mistakes (and strengths)

z Personal data are not disclosed

z Knowledge can be shared with other reviewers and prospective thesis writers

KNOWLEDGE SHARING PROCESSES

z Reviewing knowledge represented in database

z Primary purpose: generate review report

z Slides as „spin-off“

Örepresent re-usable knowledge Öcan be fed back into the process

advisor author reviewer

database thesis

slides

good practice / common mistakes

form review

report feedback loop

advisor

advisor authorauthor reviewerreviewer

database thesis

slides

good practice / common mistakes

form review

report feedback loop

RESULT AND OUTLOOK

z Technique described in use by the author for several years

z Especially useful for “intermediate theses” (“Große Studienarbeiten”)

ÖGenerates teaching material for advisory seminar

z Further development

ÖGradation support, similar to PDF form

ÖSeparation of instance-specific and generalized information in comments

ÖThesis reviewing as a learning system:

Repeated usage of the system leads to an

accommodation of the hierarchy of categories, criteria and generalized comments.

DEMO