ASPASIA Knowledge Systems - Decision Supported Information Retrieval and Knowledge Access

One of the important components respectively modules of the system is the Decision Support Unit (DSU) realized as a set of sub modules. The main target of this unit is to support and assist the user in accessing knowledge objects according to the specified task demands and formulated requirements.

The sub modules are designed in different manner to cover (and combine) different ways of information retrieval and provide accordingly a more wider spectrum of retrieval tools as realized in conventional systems. The aspect is here the rule baesd combination of structured and unstructured information (raw data).The user decides himself, which kind of tool (or combined tools) he would like to use. The choice is either related to his own "research behavior" or the degree of a-priori-knowledge concerning the definition of his request task is a measure for choosing the right tool or the right combination of tools.

In practice, to warrant a wide range of accessing facilities, a combination of the sub modules might be used.

Fig. 1: Submodules and overlapping sets of knowledge objects

Corresponding to the chosen tools resp. sub modules different sets of conceptually overlapping sets of knowledge objects are recommended to the user. The Decision Support Unit is divided into the 4 sub modules (tools):

According to the questionnaire based task formulation specified by the user a set of objects are recommended. The user is supported and assisted by multi loop feedback processes, either in optimising his information request or in evaluating recommended objects in relation to his given task. This module demands a more distinctive a-priori-knowledge of the user with respect to his formulated task and request target (see CBR in Backgrounds).

The main approach of the DSU consists in a dynamically relationship between subject trees (SCT) and task trees (TCT), whereby the TCT is structured in the same way as the SCT based on centroids as tree nodes. The correspondence between SCT and TCT is realized as a structured relationship function SRF, which implies similarity terms and coefficients, feedback terms, calculated out of the feedback data given by the user and further coefficients, calculated out of the statistical analysis of the system history data like the accumulated satisfying factor (ASF, TC, SF). The dynamically growing and expansion of the TCT is referenced to the SCT by the varying relationship function SRF - the TCT is organically mapped with the SCT. Each object in the TCT is formalized as a Structured Task Description Vector - STDV.

A permanently running module, the relationship module, calculates the relationship function out of the data given within questionnaires, filled in by the user, similarity estimations using the centroid vectors and further statistical history data calculations.

Internal feedback loops based on fitness functions evaluate and determine which parts of the user specified task description are fitting well (and in which degree) to the objects recommended by the system - the user is assisted and supported in optimizing his task description. This fitness function also gives implicitly advises to the user to avoid generation of inconsistencies and in-dicates probable contradictions. A further internal loop considers statistical history data to facilitate modifications of the task description by the user.