How do the Det works?

Det is interactive expert system that gets conclusions from questions. From the side of end-user it is as simple as hypertext. However Det is not limited by preconceived links, as all hypertext systems are. Choosing the next question Det takes into account what and how have you answered before.

Det supports the following data structures:

Det compiler automatically converts these initially various data structures into one large rule set. This is why Det can work with many decision trees at once. This system is remotely related to the rule-based forward reasoning expert systems. However it has many specific features, created on the base of authors practical experience in determining biological objects.

Many rule-based expert systems start from the empty conclusion set, step by step accepting the new intermediate and then final conclusions. Det starts with the full conclusion set, containing all possible conclusions, defined in the database. After each question, it rejects a certain part of possible conclusions until only one conclusion remains.

Each Det rule contains the following information:

If the user chooses the first proposition, then all conclusions from the list of the alternative proposition are rejected. If the user chooses the alternative proposition, then all conclusions from the list of the first proposition are rejected. From Det 3.001 it is possible to define certain important exceptions logic setup dialog.

Dichotomic keys

Far before creating computers botanists, entomologists and other scientists wrote the hypertext-like keys to determine biological objects. The main element in these keys is a question. It has the entry point (situation identifier), and two propositions, from that only one can be true in a certain situation. Each proposition is followed by number of new question or short final conclusion. Assuming "yes" for a certain proposition, you go from question to question until some conclusion if reached. For example:

15 The lichen is brightly yellow. 16
-- The lichen has different colour, only sometimes it is a little yellowish 20;
16 The lichen has a round shape with lobes that are flat (1-2 mm). Typically it is rather large. There are numerous orange-yellow apothecia (fruit bodies). If you drop 10 % KOH solution on it, the bright purple colour appear. - Xantoria parietina
-- The lichen is usually small and don't change colour after KOH treatment' 17;
17 Thallus remembers small scales, only 2-3 mm length, sulphur colour. These scales makes rosette. As a rule, this lichen grows on deciduous trees - Candelaria sp
-- Thallus quite large, up to 50 mm in diameter, laciniate, lemon colour. As a rule, grows on pines and other coniferous trees, sometimes on stones. - Cetraria pinastri;

To use in Det, such old manuals require only minor adaptation. Det supports nested keys, very common in old literature. In addition, Det can use several keys at the same time, switching where necessary from one to another. This is important, because as a rule it is possible to write a large number of different keys for the same objects. Det finds the common internal points, where it is possible to jump from one key to another, not restarting from beginning. Hence from several old keys, written by different authors, you can create the interactive system, much more powerful than any of these keys, taken separately.

Description tables

While the dichotomic keys are optimal for using earlier written manuals, the description tables is the best tool to create the expert system from new data. For example, before writing a key, many scientists draw tables like this:

  The lichen is yellow Lichen is quite large Red colour after KOH treatment
Xantoria parietina yes yes yes
Candelaria sp no no no
Cetraria pinastri yes can vary no

In reality they are obviously much larger. Det format is a little different from the simple table about, but translation is usually straightforward. Situations, when some questions cannot be answered for a certain object are understood and handled.

Tables are easier to write than keys, and more effective. This is because from one table it is possible to generate sometimes very large number of possible trees. Despite some expert systems do generate trees from tables, Det does not. It uses tables 'as is', escaping memory overflow when the number of possible trees exceeds all clever limits.

Rules

Simple rules can be called an "assembly language" of Det. They can be used to take use of small pieces of knowledge, not sufficient to build the whole key or table. For example, it is possible simply to state: if the lichen is yellow, it is Xantoria parietina. This is obviously not true, but in some cases one very characteristic feature really can help to identify the species. With rules, it is possible to perform some additional tricks, for example, launch new determining process after reaching certain final conclusion.

Final descriptions

A long description can appear after the final conclusion is reached. There is no difference, which of the defined keys or other data structures were used to reach the condition. Hence the description need not be repeated in each key. The program layout temporary changes to display the description more effectively.

Images

Det follows tradition to provide maximum information for each question. Independently the data structure being used, each question has two propositions (main and alternative), and each of them can have separate illustration. Illustrations are kept in .jpg format, ideal for true-colour real world images, including photos and microphotos.

"Connect" button

For each information unit, it is possible to specify certain Internet address. Det has special button ("Connect") that launches default Internet browser to open the address, specified for the current rule. This button is useful when the objects being determined are associated with certain Internet addresses. In addition, it could be much easier for you to get permission to include one or another image into your system if you promise to provide corresponding link to the owners homepage.

Det also supports national languages.

With all these features, Det is a serious tool to build expert systems for recognition of different objects. If some additional features are desired, it is always possible to discuss the idea about specialised Det builds.

Now Lithuania participate in the Fifth European Framework Programme and other European funds with the rights of a member of European Union. On the base of this, Vilnius Gediminas Technical University can write the application (CRAFT or RTD) to create a specialised expert system for several small or medium size enterprises. We also have scientific base on bioengineering and many other areas.

Home Detailed description of Det language Example Contact us