Data Mining to Discover a New Lighting MaterialApplication Note Zaptron Systems, Inc.
Goal:
Discover new materials that have the same properties of some existing materials
without patent infringement.
Objective:
Discover a new rear earth-containing phosphor for florescent lighting from the
published data of 46 related phosphor patents.
Features used:
[Ce] -- Ce content | |
[B] -- B content | |
[Mg] -- Mg content | |
[Tb] -- Tb content | |
(Z/r) -- Average atomic parameter | |
X -- Electronegativity |
Method:
Data were exptropolated to the outside of the patent scope to obtain a new material that has the exact same property as some existing ones. In particular, our data mining has discovered the relationship between the material properties and its composition, by separating data into a subspace where a mathematical model is built. From this model, it was easily found out that if the Mg-containing material is replaced by a Zn-containing material in a proportional mammer, a new phosphor could be discovered.
Result:
A better phosphor with higher brightness and energy saving feature has been
discovered and manufactured. This example shows that new material patents can be
development by mining the patent data from history without patent infringement.
© Copyright, 1998-2000, Zaptron Systems, Inc. Back to Home.