Branches of Artificial Intelligence?

AI is basically a collection of concepts called the branches. Stated below are some of the branches. It is rather apparent that all branches have not been included, one reason being, some of them are yet to be identified.

Logical AI

What a computer knows about the world in general, what it knows about a specific situation in which it has to act its goals are all stored in some mathematical or logical language. The program decides what action to take that is appropriate for achieving its goal by referring these stored data. In practice this is what happens in a normal PC - the machine depending on the request takes an action depending on the program. But the complexity, ambiguity and volume of data associated with AI in this context is far more ahead of a normal PC running a software.

Search

AI programs are often required to examine large number of possibilities such as what a chess game does. Discoveries are consistently being made in this branch to do this more efficiently.

Pattern Recognition

A machine program to compare an observation made by it with a standard set of patterns. This is actually what happens in voice recognition systems and vision recognition systems, which have reasonably made their advent into the commercial world. More complex comparison techniques which require more complex logic and vigorous programming are in the making now.

Representation of Facts/Knowledge

Perhaps this is one of the branches that is still at its early stages and requires lot of development. Since knowledge is available in many forms, the method of representation must be such that it can be easily used in reasoning, it can be easily examined and updated, it can be easily judged as relevant or irrelevant to when applied to particular problems. The issue of representation of facts remains an obscure one to the scientists involved in the development of AI since facts are voluminous and are constantly changing. Some current methods include constraints, frames, logic, relational databases, scripts, semantic nets and summary charts which hardly are satisfactory for the effective implementation of AI. At present languages of mathematical logic are used for this purpose.

Inference

Inference is the process of creating explicit representations of knowledge from implicit ones. It can be regarded as the creation of knowledge. Deductive inference uses a set of axioms to produce new statements. Inductive inference starts with a set of facts or observations and produces generalisations, descriptions and laws. Although mathematical logical deduction is adequate for some purposes, new methods of non-monotonic inference have been developed since the 1970's.

Common Sense (Knowledge and Reasoning)

This is one area that might help people who argue against intelligent computers. This is where AI is farthest from human-level. In spite of the fact that there is reasonable progress like the development of non-monotonic reasoning and theories of action, yet more range of ideas is pending for satisfactory implementation.

Planning

Planning programs generally start with the general facts about the effects of action, facts about the particular situation and a statement of goal. Then a strategy (a sequence of actions in most cases) is built to achieve this goal.

Epistemology

This is the study of the kinds of knowledge that are required for solving problems in the world.

Ontology

This is the study of the kinds of things that exist in the physical world. In AI programs deal a lot with various kinds of objects and we study what these kinds are and their basic properties. Emphasis on Ontology started just at the beginning of the 1990's.

Heuristics

A heuristic is a way of trying to discover an idea or something else imbedded in a program. Heuristic functions are used in certain approaches to search or measure how far the program is from reaching its goal.