Functionality

DEXiWin supports the following tasks:

  1. Developing DEX models: creating and editing DEX models and their components:

    • Attributes: variables that represent decision subproblems and properties, observed at decision alternatives.

    • Tree of attributes: a hierarchical structure representing the decomposition of the decision problem.

    • Scales: discrete and continuous scales that define admissible values that can be assigned to attributes.

    • Aggregation and discretization functions: functions that define the aggregation/discretization of attributes from the bottom to top of the attribute tree.

  2. Evaluation and analysis of decision alternatives:

    • Creation and editing of alternatives and their values, assigned to individual attributes.

    • Evaluation of alternatives: a bottom up aggregation of alternatives’ values in accordance with model structure and defined functions.

    • Analysis of alternatives: what-if analysis, selective explanation, comparison of alternatives, “plus-minus” analysis, target analysis.

  3. Other graphical-user-interface and input-output features:

    • Loading and saving DEX models in the native XML format (.dxi files).

    • Copying, pasting, importing and exporting DEX models and their parts in various formats: Json, tab-delimited and CSV (Comma-Separated Value).

    • Reporting: creating, combining, displaying and exporting various reports.

    • Charts: creating and displaying various customizable charts.

    • Tree graphic view: graphic display of model structure.

Compatibility

DEXiWin is backward compatible with DEXi. It implements all DEXi’s features, except loading and saving DEXi models in obsolete data formats (.dax and pre-2000 .xml). DEXiWin can read and process .dxi files produced by DEXi. Also, unless new DEXiWin features are used, .dxi files remain readable by DEXi.

In comparison with DEXi, DEXiWin introduces several new features:

  • DEXiWin supports all DEX method extensions introduced in DEXi Suite.

  • Additional features for handling decision rules: considering symmetricity, marginal values, representation using decision trees, graphical interpretation using linear functions, multilinear interpolation and QQ evaluation.

  • “Plus-minus-1” analysis is generalized to “Plus-Minus-AnyValue” analysis.

  • Added Qualitative-Quantitative (QQ) evaluation for ranking alternatives within qualitative classes.

  • Added new report types, supporting a flexible composition of reports.

  • Added new chart types. Some of the old ones have been redesigned and extended, too.

  • Implemented many small, but important improvements in the user interface, for instance displaying model components in user-selectable columns, tree-structured display/editing of evaluation results, consistent use of colors for displaying ‘good’ and ‘bad’ value categories, predefined scales, etc.

  • Incorporated the functionality of DEXiTree.

  • Most user-defined settings are now remembered in .dxi files.