Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Uncovering Implicit Rules in Medicine Diagnosis
Blekinge Institute of Technology, Faculty of Computing, Department of Software Engineering.ORCID iD: 0000-0003-0396-1993
Blekinge Institute of Technology, Faculty of Computing, Department of Software Engineering.
(English)Manuscript (preprint) (Other academic)
Abstract [en]

Context:  Decisions taken by experts may be based on explicit and implicit rules. By uncovering the implicit rules the expert may have the possibility to explain its decisions in a better way, both for itself and the person which the decision is affecting. In the area of medicine, laws are enforcing the expert to be able to explain its decision when a patient is complaining about a decision. Another vital aspect is the ability of the expert to explain to the patient why a certain decision is taken, and the risks associated with the decision.

Objective: To investigate if it is possible for a machine learning pipe-line to find implicit rules used by experts, when they decide if a patient could be operated or not.

Method: We conduct an analysis of a data set, containing information about patients and the decision if an operation should be performed or not.

Results: We have implemented a machine learning pipe-line which supports detection of implicit rules in a data set. The detection of the implicit rules are supported by an algorithm which implements an agglomerative merging of feature values. We have improved the original algorithm by showing the boarders of the feature values of a discretization bin.

Conclusion: The analysis of the data set shows it is possible to find implicit rules used by the experts with the help of an agglomerative merging of feature values.

Keywords [en]
agglomerative feature value merging; implicit rules detection; decision tree
National Category
Computer Sciences
Identifiers
URN: urn:nbn:se:bth-15179OAI: oai:DiVA.org:bth-15179DiVA, id: diva2:1143700
Available from: 2017-09-22 Created: 2017-09-22 Last updated: 2022-11-23Bibliographically approved
In thesis
1. Towards Intent-Driven Systems
Open this publication in new window or tab >>Towards Intent-Driven Systems
2017 (English)Licentiate thesis, comprehensive summary (Other academic)
Abstract [en]

Context: Software supporting an enterprise’s business, also known as a business support system, needs to support the correlation of activities between actors as well as influence the activities based on knowledge about the value networks in which the enterprise acts. This can be supported with the help of intent-driven systems. The aim of intent-driven systems is to capture stakeholders’ intents and transform these into a form that enables computer processing of them. Only then are different machine actors able to negotiate with each other on behalf of their respective stakeholders and their intents, and suggest a mutually beneficial agreement.

Objective: When building a business support system it is critical to separate the business model of the business support system itself from the business models used by the enterprise which is using the business support system. The core idea of intent-driven systems is the possibility to change behavior of the system itself, based on stakeholder intents. This requires a separation of concerns between the parts of the system used to execute the stakeholder business, and the parts which are used to design the business based on stakeholder intents. The business studio is a software that supports the realization of business models used by the enterprise by configuring the capabilities provided by the business support system. The aim is to find out how we can support the design of a business studio which is based on intent-driven systems.

Method: We are using the design science framework as our research frame- work. During our design science study we have used the following research methods: systematic literature review, case study, quasi experiment, and action research.

Results: We have produced two design artifacts as a start to be able to support the design of a business studio. These artifacts are the models and quasi-experiment in Chapter 3, and the action research in Chapter 4. The models found during the case study have proved to be a valuable artifact for the stakeholder. The results from the quasi-experiment and the action research are seen as new problem solving knowledge by the stakeholder.

Conclusion: The synthesis shows a need for further research regarding semantic interchange of information, actor interaction in intent-driven systems, and the governance of intent-driven systems.

Place, publisher, year, edition, pages
Karlskrona: Blekinge Tekniska Högskola, 2017. p. 129
Series
Blekinge Institute of Technology Licentiate Dissertation Series, ISSN 1650-2140 ; 1
Keywords
business intent, business support system, intent-driven system, compositional system
National Category
Software Engineering
Identifiers
urn:nbn:se:bth-15141 (URN)978-91-7295-342-0 (ISBN)
Opponent
Supervisors
Projects
Professional Licentiate of Engineering Research School
Funder
Knowledge Foundation
Available from: 2017-09-22 Created: 2017-09-15 Last updated: 2018-01-13Bibliographically approved

Open Access in DiVA

No full text in DiVA

Authority records

Svahnberg, Mikael

Search in DiVA

By author/editor
Silvander, JohanSvahnberg, Mikael
By organisation
Department of Software Engineering
Computer Sciences

Search outside of DiVA

GoogleGoogle Scholar

urn-nbn

Altmetric score

urn-nbn
Total: 122 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf