The Doctus knowledge-based expert system shell is a software for modelling expert knowledge which we developed and used mostly to support the decision taker.

The Doctus development team consists of Zoltán, Jolán, Zoltán, and me.

Download an evaluation copy: Doctus 3.0

See the documentation: onine or download

Deduction (Rule-Based Reasoning, RBR)

In deductive or rule-based reasoning the rules of the decisions are explicitly articulated and entered into Doctus. By applying these rules to the decision alternatives we get the evaluation of the alternatives. This approach is typically used in original decision situations; i.e. when the experts do not have sufficient experience for case-based reasoning. (read more)

Induction (Case-Based Reasoning, CBR)

In induction the experts describe cases from their experience, including the outcomes of these cases. The machine learning algorithm of Doctus infers the rules from the provided case set. This type of reasoning requires a large number of cases, therefore it can be used in routine decision situations. (read more)

Reduction (Extracted Rule-Based Reasoning, ERBR)

Using induction Doctus generates rules that describe the cases of experience. These rules, however, usually use significantly fewer attributes than how many were used to describe these cases. From such case-based knowledge base Doctus can extract a new rule-based one, which is why we call this type of reasoning reduction: we are reducing the number of attributes. We are learning from the cases of experience. (read more)

Illustration

Sample knowledge bases: Tender (deduction), Ability (induction)

Quick decision: Employee evaluation, Management, Outsourcing, Subsidiaries

Knowledge-based corporate portal: Doctus Executive Portal demo

Knowledge-based collaboration: Doctus Knowledge Factory (video & demo)

Knowledge export: web page with reasoning
Knowledge import: data mining

Publications

We have written papers for various journals, gave presentations at conferences and ran workshops about and with Doctus. Some of our publications are conceptual, on developing Doctus and our ideas behind them, while some are real-life applications typically in supporting the decision taker. We list some of these paper here, the rest you can find in the papers section.

Conceptual papers:

Baracskai-Dörfler-Velencei (2008) Concept Mapping and Expert Systems: Exploring Synergies, CMC 2008, 22-25 September 2008, Tallinn, Estonia & Helsinki, Finland: Vol. 3: 70-74
Baracskai-Dörfler-Velencei (2007) The ES Could Probably Know More - But Man Would Not Make Better Business Decisions, VIPSI 2007 Slovenia, 8-11 October 2007, Bled, Slovenia
Baracskai-Velencei-Dörfler (2005) Reductive Reasoning, MNJE, 1(1): 59-66
Baracskai-Velencei-Dörfler (2005) MABE Methodological Framework, ECRM 2005, 21-22 April 2005, Paris, France: 21-29

Applications:

Baracskai-Dörfler-Kis-Dózsai-Velencei (2008) Supporting the Decision Taker: by Coaching and Expert System, AoM 2008, 8-13 August 2008, Anaheim, CA
Baracskai-Dörfler-Velencei (2007) Ocjena rizika nije upravljanje rizikom, SM, 12(1-2): 15-17 [published in Croatian]
Baracskai-Velencei-Dörfler (2005) Knowledge Visualization by Doctus Knowledge Galaxy, IPS-USA-2005, 7-10 July 2005, Cambridge, MA
Baracskai-Velencei-Dörfler (2003) Knowledge Sharing by Knowledge Factory, ICC&IE 2003, 2-4 February 2003, San Francisco, CA: 64-68
Baracskai-Dörfler (2002) Intelligent Business Decisions by eFLOW Portal, iiWAS 2002, 10-12 September 2002, Bandung, Indonesia: 218-224
Baracskai-Chikán-Dörfler-Velencei (2001) From Hard Data to Soft Decision, ICC&IE 2001, 1-3 November 2001, Montreal, QC: 632-637

Future Development

Here we will outline our ideas for developing Doctus and the intelligent applications based on it - regardless whether we will actually engage in implementing these or not. If we have something in development we will also provide a sneak-peak.