|
iDEA: Drexel E-repository and Archives >
Drexel Academic Community >
College of Information Science and Technology >
Research Day Posters (IST) >
A carrot for knowledge management: Automatic generation of draft reports as a benefit to encourage contribution to KM systems
Please use this identifier to cite or link to this item:
http://hdl.handle.net/1860/1590
|
| Title: | A carrot for knowledge management: Automatic generation of draft reports as a benefit to encourage contribution to KM systems |
| Authors: | Gunawardena, Sidath
Weber, Rosina O. |
| Keywords: | Case representation
Knowledge artifacts
Knowledge management |
| Issue Date: | 17-Apr-2007 |
| Publisher: | Drexel University. College of Information Science and Technology. |
| Series/Report no.: | IST Research Day 2007 posters |
| Abstract: | Many Knowledge Management (KM) systems underperform due to lack of user
contribution. Users who do not see benefits commensurate to the effort it takes to
contribute to the KM system will not fully participate. How to encourage users to share
knowledge is a key issue of KM. In many organizational environments, report generation
is a regular and burdensome task. Where a KM system exists, this is a duplication of
work. We propose a methodology that shows the user a tangible benefit of participating
in the KM system: a lessening of the burden of another onerous task, that of report
generation. We applied this methodology in a deployed KM system built to cater to the
needs of a distributed scientific community. The scientists contribute knowledge artifacts
about their research to the KM system in the form of Learning Units (LUs). We use a
representation for the fields of the LU such that it leads the user to enter their findings in
a structured manner. Associations between LUs’s fields create the overall flow for a
report. To generate the draft reports we use template-based natural language generation.
Selecting which template to apply is determined by rules generated from examining a
subset of units in the system. The resulting drafts still require fine tuning. The user has to
address issues such as tense consistency, redundancy, and incompleteness. However, our
goal is to relieve the user of the effort required to gather and structure the information; to
provide a draft, not the end product. |
| URI: | http://hdl.handle.net/1860/1590 |
| Appears in Collections: | Research Day Posters (IST)
|
Items in iDEA are protected by copyright, with all rights reserved, unless otherwise indicated.
|
www.drexel.edu
