HOME RESEARCH AREA IDEAS METHOD FINDINGS

Predicting Future Technology Trajectories: The Internet as a Case Study.

A UCLA Internet Research Initiative Project.

Technology:

An application of knowledge

Internet

Lighting

Transport


United States Patent 10,681,830
Goodenough June 9, 2020

Sensory device protector

Abstract

The present invention provides sensory device combinations, such as a portable touch screen smart phone, which possesses unexpectedly superior protection against impact damage. The sensory device...


Inventors: Goodenough; Troy (Mindoro, WI)
Applicant:
Name City State Country Type

Goodenough; Troy

Mindoro

WI

US
Family ID: 1000002748060
Appl. No.: 15/731,483
Filed: June 19, 2017
Related U.S. Patent Documents
Application Number Filing Date Patent Number Issue Date
62493005 Jun 20, 2016

Current U.S. Class: 1/1
Current CPC Class: H05K 5/03 (20130101); G06F 3/0416 (20130101); G06F 3/044 (20130101)
Current International Class: G06F 3/044 (20060101); G06F 3/041 (20060101); H05K 5/03 (20060101)

Research Area

Using the internet as a case study, this study introduces a data-driven method to predict technology expansion trajectories based on the knowledge relatedness between patent classes calculated using patent data from the United States Patent and Trademark Office (USPTO) and the Cooperative Patent Classification (CPC) system.

Ideas

  • Use of knowledge relatedness as a determinant for the expansion trajectories of a certain technology
  • Use of the Internet as a test case for the prediction method due to high volumes of Internet patenting activity

Method

  • Calculation of knowledge relatedness between spheres of knowledge for 1980 to 2018
  • Retrieval of internet patents and their associated patent classes using keyword search
  • Assigning predictions as spheres of knowledge with knowledge relatedness (relative to those associated to internet patents) higher than a predetermined threshold
  • Testing the prediction accuracy and optimal prediction thresholds using the ROC curve

Findings

  • There is an overall positive correlation between knowledge relatedness and the predictability of expansion trajectories
  • The mean optimal prediction threshold is at a knowledge relatedness of 0.22
  • The prediction method has an average accuracy of 74% between 1980 and 2004
  • Accuracy drops drastically past 2005, likely as a result of a critical saturation of internet technology in the knowledge space

Knowledge relatedness Matrix

A21 ... H04 H05 H06 ... Y04
A21 0.33 ... 0.27 0.15 0.01 ... 0.20
: ... ... ... ... ... ... ...
G05 0.11 ... 0.08 0.12 0.23 ... 0.01
G06 0 ... 0.04 0.1 0 ... 0
G07 0 ... 0.30 0.13 0 ... 0.07
: ... ... ... ... ... ... ...
Y04 0 ... 0 0.50 0.25 ... 0


Knowledge Relatedness

Knowledge RelatednessH05G06 = Co-Classification CountH05H06 / √(Total CountH05 * Total CountG06)

Knowledge Relatedness Network (1980)

Prediction Trajectories on Knowledge Relatedness Network (1980)

Conclusion

For more details, see https://tech-predict.humspace.ucla.edu/

Research Area


The spread of new ideas and technology is often a seemingly unpredictable phenomenon. One minute, a technology is a novelty that only a few enthusiasts talk about; the next minute, it explodes into something that even the layman knows by heart. The rapidity at which a technology is adopted has only intensified with the increasing interconnectedness among people, companies and governments worldwide as a result of globalization. This unpredictability is exacerbated with uncertainty over which spheres of knowledge a technology is likely to disrupt, where a sphere of knowledge refers to a collection of knowledge regarding an area of expertise, while a technology refers to an application of knowledge. The relationship between knowledge spheres and technology is explicit; different spheres of knowledge contribute directly to the development of technologies, which in itself are applications of its related knowledge spheres. An understanding of the impact that the diffusion of a technology may have on knowledge spheres can make or break the decisions made by key stakeholders of economies, namely, innovators, businessmen, investors and policymakers.

The significance of the impact of technological diffusion on knowledge spheres is clear for innovators. Knowledge spheres represent domains of innovation for innovators. By understanding technological expansion trajectories, innovators can gain from the suggestion of new design opportunities and directions. For businessmen and investors, knowledge spheres translate into types of industries. An accurate identification of an industry that is likely to experience a disruption from an emerging technology can allow a businessman to adapt his operations accordingly and aid an investor in allocating his assets, both of which will allow these stakeholders to reap enormous benefits. Conversely, a businessman or investor who fails to foresee a disruption of an industry by an emerging technology may fail to react in time during its onset. In turn, a business may find themselves losing their competitive edge, or worse, may eventually be forced out of business. It is also no surprise that technological change can make or break an entire economy. Given the right government policies to steer an economy in the right direction to adopt a certain technology, a country can experience an unprecedented surge in economic growth that may set the path for further development. A misjudgment of policies could also lead an economy in the opposite direction of economic growth.

It is arguably unthinkable that these decisions, which are based on nothing more than intuition and luck, could mean the difference between thriving and shutting down for a company and between an economic boom and possibly an economic downturn for a country. It would be greatly instrumental if a predictor for the pace and pattern of technological adoption were available to reduce the level of uncertainty in the decision-making process. Today, we are anticipating the same rapidity and uncertainty regarding the technological adoption of emerging technologies such as artificial intelligence, internet of things and blockchain. Can a prediction method for how, when and where these technologies will spread be developed to aid innovators, investors, businessmen and policymakers?

Ideas


Knowledge Space and Technological Relatedness

While a seemingly intangible subject, patent data allows for an in-depth study of knowledge and technology. Patents are intellectual property rights granted to an innovator over a unique invention, and can be viewed as a tangible materialization of knowledge. Patents are further categorized into patent technology classes according to patent classification systems, such as the Cooperative Patent Classification (CPC) system. Each of these classes represent a sphere of knowledge, which can range from apparel to electric heating. Due to its availability and accessibility, patent data has been increasingly used by researchers to study knowledge and technology.

Fig. 1. US knowledge space. (a) 1975, (b) 1985, (c) 1995 and (d) 2005.
Notes: Black = Chemicals (1), Green = Computers = Communications (2), Yellow = Drugs & Medical (3), Red = Electronics (4), Blue = Mechanical (5), Grey = Miscellaneous (6).


In their paper titled ""Mapping knowledge space and technological relatedness in US cities,” Kogler, Rigby and Tucker create a measure of technological relatedness using patent data from the United States Patent and Trademark Office (USPTO). The measure of technological relatedness is an index for how “close” in proximity two patent classes are, and is determined by the number of patents that are co-classified within the patent classes. By determining the technological relatedness between every pair of patent classes out of the 438 primary patent classes, they were able to produce a map of the US knowledge space at different periods of time as seen in Figure 1.

They go on to complement the US knowledge space and the technological relatedness data with data on inventor locations to determine the relative competitive advantage that each city in the US possesses, thus making a case for the spread of innovation based on the relative competitive advantage of a specific geographical location.

This demonstrates the possibility of mapping out the relatedness of two spheres of knowledge to build an overall network of the spheres of knowledge, where each sphere is connected to another by their established knowledge relatedness. In doing so, a knowledge space is created.


Analysis of Expansion Trajectories

Fig. 2. Total technology space map overlaid with the HEV design knowledge base and its most likely expansion paths. The map includes the IPC class ID, title and entered year of each HEV’s historical technology entry

In "
Overlay technology space map for analyzing design knowledge base of a technology domain: the case of hybrid electric vehicles," Song et al. takes the knowledge space one step further and use it to analyze the expansion trajectories of technologies into spheres of knowledge beyond the present spheres of knowledge they belong in. Their paper introduces a “network-based methodology for visualizing and analyzing the structure and expansion trajectories of the design knowledge base of a given technology domain.” Using hybrid electric vehicles (HEVs) as the technology domain to be studied, they first retrieved the set of patents related to HEVs and determined the patent classes that those patents fell under. Those patent classes would make up the “design knowledge base” of the HEV domain. Taking the patent classes in the design knowledge base, they proceeded to highlight them on the overall knowledge space map similar to the one created in Kogler, Rigby and Tucker’s research. Thereafter, they identified the strongest links from those highlighted spheres of knowledge to other spheres of knowledge on the knowledge space, as can be seen in Figure 2. Those strongest links together hence suggested the most likely paths through which the technology would have expanded from earlier-covered to later-covered spheres of knowledge.

According to Song et al., In particular, those unexplored spheres of knowledge that have the highest relatedness to the established design knowledge base of a domain may present most feasible opportunities for technological expansion, because the high innate knowledge relatedness would enable the ease for innovators in the domain to comprehend these unexplored spheres of knowledge.


Research Questions and Hypothesis

This research project will recreate the overlay technology space map used by Song et al. and test its accuracy as a prediction method for technological expansion trajectories. Internet technology will be used as the case study for this prediction method, due to the extensiveness and availability of patents pertaining to the Internet. It will answer the following questions:

Is the technological relatedness of a pair of knowledge spheres positively related to the predictability of a technology that appears only in one of the pair of knowledge spheres?

If so, can this knowledge be used to develop a practical method of prediction for future technology trajectories?

Research Methodology


Some text about me. Some text about me. I am lorem ipsum consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat.

Findings


Patent Counts


Internet Patents CPC Subsection Distribution


Technology Relatedness Network Graph


Some text about me. Some text about me. I am lorem ipsum consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat.

Technology Relatedness Matrix Aggregate Measures


AUC Graphs 1980-2018


Mean AUC