WEB information correlativity computing algorithm

Abstract

Disclosed is a method used for computing the WEB information relativity and the internet tree structure connection; the method includes the computation of node rank and the weight computation of the relation; wherein, 1) the node rank=n1* the node access number^m1+n2*(the node access number / (the referred number of the node +b))^m2+n3* the referred number of the node^m3; 2) W=W`*attenuation coefficient lambda+(U-U`); wherein, the U is the computation weight value of the day and has no relation with the prior weight value; the computation formula of U is as follows: U=n1*the number of accessing B from A^m1+n2*((the access number of B-the number of accessing B from A) / (the referred number of B+ b))^m2+n4* the rank of B^m4; A and B are two nodes in the reticular structure; the U` is the former computation weight value; the W is the current valid weight value; the W` is the former valid weight value; the formula of W indicates that the current valid weight value equals to the sum of the net value of the computation weight value and the partial inheritance of the former valid weight value; the computation weight value of the day relates to the number of accessing B from A, the access number of B, the referred number of B and the rank of B.

Description

technical field [0001] The invention relates to a calculation method of WEB information relativity in semantic network, especially a method for calculating WEB information relativity used for Internet tree structure connection. Background technique [0002] First of all, let me introduce that the Markov process with discrete time and state is called a Markov chain, denoted as {Xn=X(n), n=0, 1, 2,...}, in the time set T1={0 , 1, 2, ...}, the result of successive observations on the process of discrete states, the state space of the chain is recorded as I={a1, a2, ...}, ai∈R. The conditional probability Pij(m, m+ n)=P{Xm+n=aj|Xm=ai} is the transition probability that the Markov chain is in state ai at time m+n and transitions to state aj at time m+n. [0003] PageRank (webpage level), which is a system for webpage rating developed by Google founders Larry Page and Sergey Brin at Stanford University, webpage level is the basis of all Google web search tools. As an organizatio...

AI Technical Summary

Problems solved by technology

[0006] However, at present, in traditional Internet applications, information is basically tree-like, and there are only limited hierarchical relationships (<5)

This kind of Internet application model is not suitable for calculating the relationship between information. The amount of calculation will increase exponentially with the increase of the number of information points.

Images