PPubMed6 Dec 2022Cited 6×

Assessing future technological impacts of patents based on the classification algorithms in machine learning: The case of electric vehicle domain Assessing future technological impacts of patents based on the classification algorithms in machine learning

Fang Han, Shengtai Zhang, Junpeng Yuan, Li Wang, Zhihong (Arry) Yao

Abstract

Early identification of the technological frontier is important for the optimal allocation of enterprises’ R&D resources and the formulation of government innovation strategies. Many scholars have used bibliometric methods to identify the technological frontier, and the citation analysis method has been widely used. However, it takes a certain amount of time to accumulate citations of patents. The existing citation analysis method cannot incorporate newly published patents, which are potentially highly cited, into the data collection of the important patents used to identify the technological

A figure from Assessing future technological impacts of patents based on the classification algorithms in machine learning: The case of electric vehicle domain Assessing future technological impacts of patents based on the classification algorithms in machine learning — fig. from the paper

§ The Valyu brief

The paper, in five minutes

Read from the full text via Valyu Answer

A machine‑learning classification framework predicts which new EV patents will become highly cited, enabling faster identification of emerging technology frontiers.

What’s new

i.Proposed a novel machine‑learning classification framework to identify technological frontiers of patents
ii.Applied the framework to the electric‑vehicle patent dataset to validate its effectiveness
iii.Demonstrated that classification models outperform traditional multiple‑regression approaches in fitting accuracy and generalization
iv.Incorporated a broader set of influencing factors, including weakly correlated citation indices, to improve prediction
v.Provided an open, extensible framework for future use of additional ML models and domains

How they did it

The authors built classification models (e.g., coarse‑grained machine‑learning algorithms) to predict patent citation frequency using a wide range of patent attributes, then compared their performance against multiple‑regression models on an electric‑vehicle patent corpus.

What they found

Classification algorithms showed higher fitting accuracy than the multiple‑regression baseline
Coarse‑grained models exhibited better generalization to unseen patents
Additional citation‑related indices displayed weak but positive correlations with citation frequency, supporting the science‑technology link

Caveats

✕Only a limited set of influencing factors (derived from literature) were included; more variables could improve predictions
✕The study evaluated a single domain (electric vehicles); results may differ in other technological fields
✕Only a few machine‑learning models were tested; exploring more algorithms could yield higher classification accuracy

§ The Valyu brief

Reading the full paper and taking notes. This takes a few seconds…

§ Ask this paper

Ask a question about this paper

Valyu reads the full text and answers from what the paper actually says.

§ Related work in other archives

Assessing future technological impacts of patents based on the classification algorithms in machine learning: The case of electric vehicle domain Assessing future technological impacts of patents based on the classification algorithms in machine learning

The paper, in five minutes

Ask a question about this paper

Similar papers in all four archives

The paper, in five minutes

Similar papers in all four archives