Brainmaker

Nanos gigantium humeris insidentes!

Introduction to Semantic Network

  • June 30, 2010 12:41 pm

Slides see http://codetopper.com/others/17/semantic-network.html

RDF(Resource Description Framework): The foundation technology of the semantic web. RDF is a highly-distributed, semantic network data model.

RDF schema(RDFS) and the Web Ontology Language(OWL): The ontology is to the semantic web as the schema is to the relational database.

Tripple-Store: the triple-store is to semantic networks waht the relational database is to the data table.

Semantic Network

  • June 30, 2010 12:27 pm

Knowledge Representation

  • June 30, 2010 11:10 am

The main methods of knowledge representation (Efraim Turban 1999) are as follows:

  • Logic.
  • Semantic Network.
  • Frames.
  • Production Rules.
  • Data Structure.
  • Algorithm.
  • Data Flow Diagram (DFD).
  • Unified Modeling Language (UML).

Knowledge Representation Using Logic

One of the oldest methods of knowledge representation is logic. It is a sub division of Philosophy. Many people have defined logic but the most common definitions are as follows:

  • A reasoning process based on science used to differentiate between correct and incorrect reasoning.

  • A process used to reason the problem to reach a conclusion or solution.

  • Method used to distinguish correct knowledge from incorrect knowledge.

Logic is further divided into two types:

  1. Propositional Logic Calculus.
  2. Predicate Logic Calculus.

Propositional Logic Calculus

Proposition means a statement is either true or false. Logic is process used to reach the conclusion. Calculus is a term used for computing. Propositional Logic used symbols for computing, such as letters of alphabets. For example, Logic gates are used for Boolean algebra:

  • NOT Gate.
  • AND Gate.
  • OR Gate.

Predicate Logic Calculus

Predicate Logic Calculus is used for AI programming. It is used to represent the knowledge in a delicate way or in a finer way. Predicate Logic concepts and rules are same as used by Propositional Logic. Predicate Programming is like the function prototype programming in C and C++ languages. Predicate is a portion of sentence containing its subjects. Declarative Programming Languages used Predicate Logic Calculus instead of Propositional Logic Calculus, such as PROLOG and LISP. A declarative program is a combination of facts and rules. Rules are relationships of facts.

Knowledge Representation Using Semantic Network

Semantic network is an easiest representation of knowledge. It is a kind of network used to represent the knowledge in graphical format. Semantic network used symbols or nodes to draw. Each node is like an arc. One of the techniques used for analysis to conceptualize the problem and verify the requirements. More close to the way human being stores the knowledge.

A Semantic NetworkA Semantic Network

wikipedia 2 Cycalike

  • June 30, 2010 7:34 am

把一篇文章A

  • (处理成简单句组成的文章B)
  • 将文章B画分出句子成分
  • (再删除不重要的句子成分)
  • 用一阶逻辑来表示一个句子,表示成知识集C
  • 文章所在的分类是micro-theory,文章名称是constant,知识集本身就是常识

这样,文章A就变成了结构化的文章C

Support Vector Machines — Different Methods

  • May 31, 2010 11:22 pm

Digested from: http://www.dtreg.com/svm.htm

A Support Vector Machine (SVM) performs classification by constructing an N-dimensional hyperplane that optimally separates the data into two categories. Before considering N-dimensional hyperplanes, let’s look at a simple 2-dimensional example. Assume we wish to perform a classification, and our data has a categorical target variable with two categories.

The simplest way to divide two groups is with a straight line, flat plane or an N-dimensional hyperplane. But what if the points are separated by a nonlinear region. Rather than fitting nonlinear curves to the data, SVM handles this by using a kernel function to map the data into a different space where a hyperplane can be used to do the separation. The concept of a kernel mapping function is very powerful. It allows SVM models to perform separations even with very complex boundaries such as shown below.


The Kernel Trick


Many kernel mapping functions can be used – probably an infinite number. But a few kernel functions have been found to work well in for a wide variety of applications. The default and recommended kernel function is the Radial Basis Function (RBF).


Kernel functions supported by DTREG: Linear: u’*v



Polynomial: (gamma*u’*v + coef0)^degree



Radial basis function: exp(-gamma*|u-v|^2)



Sigmoid (feed-forward neural network): tanh(gamma*u’*v + coef0)





Some Ideas Concerning AIBO

  • October 21, 2009 10:59 am

Since I don’t have a AIBO, I might try the following two ways to test the code:

  1. map my webcam to the certain IP
  2. use another computer to test
  3. edit the source code for local use

The best way is to try with the order 2 1 3.

Compiling Cognitive Vision

  • October 20, 2009 10:14 pm

A simply error cost me more than 7 hours.

$ cmake .

edit the CMakeCache.txt

CV_MIND_UNIT_TESTS:BOOL=OFF ON

$ make

every come across “warning as error”, remove the flag “-Werror” in the flag.make file.