student profile: Mrs Praveena Chandra


Thesis work


Supervisors: Andy DONG , Michael HARRE

Thesis abstract:

Rapid technological changes and increasingly competitive markets are pushing the need to identify successful technologies at an early stage. Investors and corporate organizations spend substantial money on forecasting various aspects of technologies that would keep them ahead of the competition. Some of these aspects include the technical performance of the technology, its commercial returns, and market acceptance. However the complex interplays of multiple factors such as the market, government policies, investments, etc. make it challenging to predict if an invention can actually be implemented into a product. In a plethora of inventions, being able to identify technologically superior inventions would be the first step towards a meaningful prediction.
Lately patent studies have been gaining interest as a forecasting tool to identify successful technologies. A patent is a set of exclusive rights granted by the government to an inventor or the assignee for a limited period of time to exclude competition, in exchange for the public disclosure of the invention. There is an vast body of research that has revealed that surface-level patent metadata of an invention, such as citation count, claims, patent life, family size, processing time and others are indicators of its value. While these techniques are useful in understanding a broad picture of the sector, they can fail to differentiate the technical feasibility of inventions that perform similar functions. Moreover it is often found that predictions based on one indicator often do not agree with the predictions made with other indicators. It is also important to note that a majority of these techniques are post hoc in their predictive ability as they use indicators that are time dependent. Valuation techniques that utilise single-level relationships such as the surface-level metadata about the patent do not account for the differences in the knowledge content between inventions. Scholars have argued that such single-level relationships, give an incomplete understanding of the knowledge flow based on which the patent value cannot be evaluated objectively. Hence the answer may lay deeper in the knowledge structure of the patent. The knowledge structure of an invention represents its relation with all the prior knowledge that it is based on. This structure crystalizes at the inception of the invention, and unlike citations, does not change with time. Using patent citation networks, this research explores the differences in knowledge structure between high value and low value inventions. More specifically, this research looks at two aspects of knowledge structures, knowledge accumulation and knowledge structure resilience.

Knowledge accumulation maybe defined as the collective body of knowledge, know-how and experiences gathered in a sector over time. Knowledge in the form of methods, procedures, experiences of success and failure come together to form a technology. This knowledge accumulates over time and forms an intricate and strongly connected network and is traceable. Since inventions have deep roots in prior inventions, this accumulated knowledge is an indicator of the value of invention. In this study resilience of knowledge structure indicates its flexibility to interruptions in knowledge flow. Interruptions in knowledge flow occur when a knowledge entity is removed from the network of an invention along with its edges thus disrupting the existing knowledge flow pathways. In which case the invention must draw the essential knowledge from elsewhere in the knowledge network for its existence. In the absence of appropriate pathways, this essential knowledge may not be able to flow to the invention thus affecting its outcome and eventually its value.

This research studies the knowledge structure of patents from three interdisciplinary sectors; inductive vibration energy harvesting, piezoelectric energy harvesting and carbon nanotubes. Using composite patent value as a proxy for technological value of the invention, this research shows that knowledge accumulation is a significant predictor of patent value. The results also suggest that the knowledge structure of high value inventions is more resilient to disruptions in knowledge flow.

Note: This profile is for a student at the University of Sydney. Views presented here are not necessarily those of the University.