Patent Information Quantifies 'Distance' Between Technologies

Nov 8, 2007
Masahide Kimura, Nikkei Electronics

Youichi Oshima, visiting professor of the Precision and Intelligence Laboratory at Tokyo Institute of Technology, explained an analytical method to quantify the scale of a target technological field and the relevance (distance) of the target technologes to semiconductor technologies based on patent information.

He made a speech at the 4th Intellectual Property Symposium of the Precision and Intelligence Laboratory at Tokyo Institute of Technology Nov 6, 2007 .

For example, if a company hopes to expand its business to a new business field with a semiconductor technology, it possibly cannot make use of the technology well. Even if the scale of the technology is large, the technologies of the target field might be too different from that of the semiconductor industry.

The analytical method quantifies this kind of situation based on patent information to help make informed decisions.

The scale of a technology in terms of patents is indicated by "SP ratio (semiconductor patent ratio)." It is calculated by dividing the number of patents that were registered in the US in a year and have a target technological keyword in the specifications by the number of patents with the keyword "semiconductor."

If an SP ratio is larger than 1, it means that there are more patents related to the keyword in other industries than in the semiconductor industry. In the 2006 data, the ratio of "display" is 1.36, and that of "automobile" is 0.89.

The relevance to semiconductor technologies is indicated by "TC distance (technology coupling distance)." This is calculated by dividing the number of patents that include "semiconductor" and a target technological term as keywords by the number of patents that have the keyword "semiconductor."

If 50% of the patents with the keyword "semiconductor" include the target keyword, the TC distance is 2. If it is 100%, the TC distance is 1. In the 2006 data, the TC distance of "computer" is 1.76, and that of "display" is 3.73.

Also, "technology gravity" was defined based on the following fact: The larger the scale of a technology or the smaller the distance between technologies, the stronger the technologies attract each other. This is calculated by the following formula: (The number of technology A's patents) × (The number of technology B's patents) / (TC distance between A and B)2.

When "technology A" is "semiconductor," the formula calculates the technology gravity toward semiconductor technologies. For example, the technology gravity of "computer" toward semiconductor technologies is very high. Recently, the technology gravities of "automobile," "cellular," "TV," "camera" and "game" are becoming higher.

In the speech, Oshima referred to a discussion of whether a company with semiconductor technologies should expand its business to the solar cell industry or the fuel cell industry.

According to that, the TC distance between solar cell and semiconductor is smaller than that between fuel cell and semiconductor, and the scale of solar cell technologies is larger than that of fuel cell technologies. Therefore, the technology gravity of solar cell is stronger than that of fuel cell.

However, the technology gravity of fuel cell has much higher growth rate than that of solar cell. In other words, the fuel cell industry has more room for research and development, and the importance of technological capabilities is higher in this industry.

Recently, MEMS (micro electro mechanical systems) technologies are used in an increasing number of fuel cell-related patents. But it is necessary to have a long-term vision because the TC distance of fuel cell is large.

On the other hand, the solar cell industry does not have ample room for technical innovations and is more suited to companies with, for example, competitive production know-how.