Create Expert Systems with VisiRule Author
VisiRule Author is a Windows desktop application which provides a graphical way for non-programmers to develop and deliver rule-based and Expert System solutions simply by drawing their decision logic.
As with Visio and MindMap, VisiRule Author is a dedicated diagramming tool which allows experts to draw a visual model of their ideas in what is often a creative and experimental exercise.
The model is in the form of a graphical decision tree flowchart.
When requested, the chart will automatically generate an interactive web-based guided interview or intelligent questionnaire to gather the necessary answers to the questions in the chart.
Business Rules and Application Logic
English-like Business Rules are created using familiar graphical icons and links. The underlying logic-based inference engine helps ensure a high degree of compliance and accuracy.
Productivity Improvements due to Graphical Programming
VisiRule transforms the experience of building decision support and business rules systems by reducing the production time while increasing the final quality of the delivered systems. Many relationships and processes are better expressed graphically.
Share Your Knowledge
VisiRule charts can be shared and discussed with colleagues with no need for to buy or install any additional software. The knowledge underpinning the business decisions is made visible in a simple and coherent manner. This helps people work collaboratively and improves group productivity without increasing costs. Teams can discuss and critique shared diagrams without requiring to go on a programming course. After all, a picture is worth a thousand words. This refers to the notion that a complex idea can be conveyed with just a single image and conveys its meaning or essence more effectively than a description does.
Simply Drag 'n Drop ... then Click 'n Link
By using a familiar decision flow-charting approach, VisiRule Author makes it easy to build systems. The author has a free hand in designing and organizing the layout of the chart. Simply drop boxes onto the canvas, size them and link them together to form the decision flow between questions.
VisiRule Question Types
VisiRule supports a wide range of question types, from simple single-choice, text entry and numerical input through to multi-choice where the user can select any number of items.
Questions are not limited to just two or three answers; they can offer as many answers as needed!
Each question has its own name, some prompt text and some explanatory guidance/links plus some associated logic tests which help 'direct the traffic' down the right branch.
VisiRule includes an interactive browser to help navigate around the chart and inspect individual items.
Statement boxes behave like computable questions, but rather than ask the user for a value, they execute some code to produce a value. Typically, this combines or analyses answers given to previous questions. For instance, you may want to calculate the density of an object given its Mass and Volume using a simple formula
You can use any of the many built-in maths or text handling routines defined in Flex or LPA Prolog, and, you can call out to programs defined in Flex and/or Prolog or any other language such as Python, C, C#.
Expressions are small bits of logic which are used to form questions and then test their answers. Expressions behave like guards, in that they have to succeed for the computation to follow that branch of reasoning.
VisiRule Author helps in the knowledge capture and automation process by intelligently constraining what can and can not done based on the its understanding of the emerging program. This reduces the scope for constructing invalid or meaningless links, improves productivity and helps spot errors early within the design process.
Modularity & Manageability
VisiRule supports multiple, linked charts and files. This means that large, complex problems can be split into smaller more manageable problems and then later combined. This also means that frequently re-occurring scenarios can be re-used in alternative situations.
Immediate Testing & Run-time Delivery
VisiRule charts can be immediately tested and executed within the Windows authoring environment, embedded within Java, C# or VB, or published directly to the internet.
You can create and link nodes in any order; define questions, answers and expressions freely. This helps you build models in the order that you want to and you can test them incrementally.
Expression Harvesting & Menu Construction
VisiRule Author harvests the values from within expressions to automatically compile the list of available menu options within single-choice and multiple-choice questions. This saves time and reduces the scope for omissions and typographical errors.
Once published, VisiRule analyses each chart and produces a dynamic online questionnaire or guided interview based on the questions and expressions used within the chart.
You can embed HTML elements in questions, explanations and conclusions such as images and video.
Export your Charts
You can export your charts as an XML table. This means you communicate everything about the chart to other applications.
Integration with other LPA AI Tools
That logic-based underpinning means that VisiRule is well placed to execute checks for consistency and completeness and produce ever more reliable and more robust decision models.
VisiRule Author is fully integrated within the Prolog IDE and can access any of the AI tools from LPA from VisiRule such as Natural Language Processing, Fuzzy Logic, Data Mining and Case-Based Reasoning.
You can use VisiRule to create Legal Expert System Applications, Financial Advisory Systems, product support and diagnostic tools, and self-assessment questionnaires. Deliver informed recommendations and advice, generate bespoke documents, custom reports, produce personal guidance, individual plans, legal checklists, letters, agreements etc.
LPA is a well established AI company and was recently recognized as such by CV Magazine.
There is scope within VisiRule for Machine Learning using feedback loops, with the charts themselves viewed as data structures by meta-programs. Work on this has recently begun and will lead to charts which adjust themselves and improve their suitability and performance over time based on performance analytics. LPA has also recently developed VisiRule FastChart which enables charts to be 'mined' from historical data using induced Decision Trees.