(2) Generalization by Meaning Weight

次へ: Minimum set of Vector 上へ: Irreducible Minimum Vector 戻る: (1)Generalization by Meaning Granularity

The attributes with small weight will become the target of generalization because they have small effects on information retrieval.

Fig.2 shows a case in which the semantic attributes with a weight smaller than 5 are generalized.

**図 2:** Generalization Methods
$\begin{figure}\begin{center} \epsfile{file=zu2.eps,height=6cm} \vspace{-5mm} \end{center}\end{figure}$

Jin'ichi Murakami 平成13年10月5日