代做Artificial Intelligence & Knowledge Engineering Assignment 3代写Python语言

Artificial Intelligence & Knowledge Engineering

laboratory

Assignment 3. A Simple knowledge base and inference.

Assignment goal

To carry out simple exercises related to knowledge representation based on facts and rules expressed in an approximately symbolic form. and inference based on Forward Chaining scheme – precisely Rete (e.g.https://en.wikipedia.org/wiki/Rete_algorithm).

Using the example of the Experta (https://experta.readthedocs.io) library (a simple Python library that implements facts and rules directly in Python by a kind of templates and macros), familiarise oneself with the possibility of using off-the-shelf solutions for representing symbolic knowledge and running inference.

Implementation

The assignment should be completed within two laboratory classes. It should be largely carried out during laboratory time, on site, and has been conceived as more of an illustration of the issue than the execution of a complete solution.

The following materials should be read carefully (and necessarily before the laboratory) to complete the tasks correctly:

●   “Lecture 4. Logic, Inference and Knowledge Representation” from ePortal,

●   Experta documentation:

○  https://experta.readthedocs.io/en/latest/index.html,

○  https://experta.readthedocs.io/en/latest/thebasics.html;

●   and examples of the simple applications of Experta:

https://github.com/nilp0inter/experta/tree/develop/docs/examples

The main theme will be the construction of a knowledge base describing a selected device or application and, based on this, construction of a simple programme supporting the user in diagnosing and solving problems. One could also think of extending the exercise with a kind of system answering simple questions about the selected device or application (patterns based on regular expressions could be used, too).

As part of your work on the exercise, you should perform.

●   Select a specific device (e.g. printer, photocopier, household appliance) or computer application and next get acquainted with its user manual, especially the section on troubleshooting.

●   On this basis, a simple semantic network of concepts relevant to the description of the device should be identified and described as facts.

●   Demonstrate capabilities of inference performed with the help of the constructed network with some well-chosen examples of queries and inference.

●   Next, extract sentences describing diagnostic rules, rules of conduct, ways of dealing with problems, etc. from the content and try to translate them into rules.

o probably the first results will still leave a lot of room for improvement, so

record the subsequent main versions and present them as a part of the exercise report as an illustration of the process of working out a solution.

●   The capabilities of the developed knowledge base should be shown on several

example problems solved by inference (minimum 3). In the selection of examples, attention should be paid to showing the possibilities of the inference process, e.g.   depth, interesting use of the rules, etc.

●   As part of the exercise report, the knowledge base code should be presented with commentary on the meaning of individual facts, rules and solutions adopted.

Evaluation

It is not expected to construct a complex expert system or to operate on complex data structures in an advanced way, although this is, of course, not prohibited.

The score for the assignment will be influenced by the following aspects:

●   the complexity of knowledge written in the form of facts and rules,

●   the complexity of the semantic web developed and the level of generalisation (e.g. defined classes of components, parts, users, etc.),

●   the complexity of the constructed rules and their multilevel combination,

●   the multi-step nature  of the inference process  and the extent to which the inference process is generalised.

Scoring:

●   Construction of a simple semantic web describing a selected device or application: 20 pts.

●   Demonstration of inference capabilities with some well-chosen examples of enquiry and simple inference: 10 pts.

●   Extraction of sentences from the manual content describing diagnostic rules, rules of conduct, how to deal with problems, etc: 10 pts.

●   Development of a knowledge base in Experta describing the knowledge expressed in these sentences: 30 pts.

●   Demonstration of the capabilities of the developed knowledge base on several example problems solved by inference (minimum 3): 30 pts.

Literature

●   “Lecture 4. Logic, Inference and Knowledge Representation” from ePortal,

●   Experta documentation:

o https://experta.readthedocs.io/en/latest/index.html,

o https://experta.readthedocs.io/en/latest/thebasics.html;

●    and examples of the simple applications of Experta:

https://github.com/nilp0inter/experta/tree/develop/docs/examples