Applying data mining techniques for technology prediction in new energy vehicle: a case study in China
TL;DR: Wang et al. as mentioned in this paper synthesized the frequent pattern growth (FP-growth) algorithm and input-output analysis to construct a new technology prediction method based on the knowledge flow perspective, takes the data of NEV patent family in 1989-2018 the Derwent patent database as a sample, divides the data according to the 5-year standard, and uses the method to identify the core and frontier technologies in the NEV field during different periods.
Abstract: Technology prediction is an important technique to help new energy vehicle (NEV) firms keep market advantage and sustainable development. Under fierce competition in the new energy industry, there is an urgent necessity for innovative technology prediction method to effectively identify core and frontier technologies for NEV firms. Among the various methods of technology prediction, one of the most frequently used methods is to make technology prediction from patent data. This paper synthesizes the frequent pattern growth (FP-growth) algorithm and input-output analysis to construct a new technology prediction method based on the knowledge flow perspective, takes the data of NEV patent family in 1989–2018 the Derwent patent database as a sample, divides the data according to the 5-year standard, and uses the method to identify the core and frontier technologies in the NEV field during different periods. Furthermore, the multiple co-occurrence method applies to analyze the technology layout and evolution patterns in China’s NEV field. The results show that the technology prediction method proposed in this paper can effectively identify core and frontier technologies to achieve NEV technology prediction.
Citations
More filters
TL;DR: This paper proposes a method that uses Variational Bayes to reduce the difference between accuracy and likelihood in text classification and proves that the proposed method within the significance level of 0.05 was more effective at calibrating the confidence than before.
Abstract: Recently, predictions based on big data have become more successful. In fact, research using images or text can make a long-imagined future come true. However, the data often contain a lot of noise, or the model does not account for the data, which increases uncertainty. Moreover, the gap between accuracy and likelihood is widening in modern predictive models. This gap may increase the uncertainty of predictions. In particular, applications such as self-driving cars and healthcare have problems that can be directly threatened by these uncertainties. Previous studies have proposed methods for reducing uncertainty in applications using images or signals. However, although studies that use natural language processing are being actively conducted, there remains insufficient discussion about uncertainty in text classification. Therefore, we propose a method that uses Variational Bayes to reduce the difference between accuracy and likelihood in text classification. This paper conducts an experiment using patent data in the field of technology management to confirm the proposed method’s practical applicability. As a result of the experiment, the calibrated confidence in the model was very small, from a minimum of 0.02 to a maximum of 0.04. Furthermore, through statistical tests, we proved that the proposed method within the significance level of 0.05 was more effective at calibrating the confidence than before.
2 citations
TL;DR: In this paper , the FP-Growth algorithm is used to analyze the range of cars spare parts for dealer car service company, which solves the problem of finding associative rules based on searching in a large volume of source data for relationships in the form of if X, then Y.
Abstract: In the current conditions of instability and a rapidly changing economy, mathematical methods and intelligent information technologies used in making managerial decisions in various fields play an important role. It is especially necessary to approach carefully the process of securing stocks of products sold, which is necessary for the profit of a car service company. The company in its activity requires a wide range of cars spare parts. The lack of necessary parts can provoke a long downtime of cars waiting for the technical maintenance or the customer's refusal from service. Excess parts that have not been sold for a long time require increased storage costs. In this article, the FP-Growth algorithm is used to analyze the range of cars spare parts for dealer car service company, which solves the problem of finding associative rules. This task is based on searching in a large volume of source data for relationships in the form of if X, then Y. The FP-Growth algorithm differs from other methods of searching for associative rules by the procedure of constructing a tree of variants of sets of objects, which allows to reduce the search for possible variations and reduce the number of iterations. To implement the proposed algorithm, the Loginom Community analytical system was used. As a result, sets of spare parts were identified, often used together in the current repair of cars.
References
More filters
TL;DR: Powell et al. as mentioned in this paper developed a network approach to organizational learning and derive firm-level, longitudinal hypotheses that link research and development alliances, experience with managing interfirm relationships, network position, rates of growth, and portfolios of collaborative activities.
Abstract: This research was supported by grants provided to the first author by the Social and Behavioral Sciences Research Institute, University of Arizona, and the Aspen Institute Nonprofit Sector Research Fund and by grants to the second author by the College of Business and Public Administration, University of Arizona. We have benefited from productive exchanges with numerous audiences to whom portions of this paper have been presented: a session at the 1994 Academy of Management meetings, the Social Organization workshop at the University of Arizona, the Work, Organizations, and Markets workshop at the Harvard Sociology Department, the 1994 SCOR Winter Conference at Stanford University, and colloquia at the business schools at the University of Alberta, UC-Berkeley, Duke, and Emory, and the JFK School at Harvard. For detailed comments on an earlier draft, we are extremely grateful to Victoria Alexander, Ashish Arora, Maryellen Kelley, Peter Marsden, Charles Kadushin, Dick Nelson, Christine Oliver, Lori Rosenkopf, Michael Sobel, Bill Starbuck, Art Stinchcombe, and anonymous reviewers at ASQ. We thank Dina Okamoto for research assistance and Linda Pike for editorial guidance. Address correspondence to Walter W. Powell, Department of Sociology, University of Arizona, Tucson, AZ 85721. We argue in this paper that when the knowledge base of an industry is both complex and expanding and the sources of expertise are widely dispersed, the locus of innovation will be found in networks of learning, rather than in individual firms. The large-scale reliance on interorganizational collaborations in the biotechnology industry reflects a fundamental and pervasive concern with access to knowledge. We develop a network approach to organizational learning and derive firm-level, longitudinal hypotheses that link research and development alliances, experience with managing interfirm relationships, network position, rates of growth, and portfolios of collaborative activities. We test these hypotheses on a sample of dedicated biotechnology firms in the years 1990-1994. Results from pooled, within-firm, time series analyses support a learning view and have broad implications for future theoretical and empirical research on organizational networks and strategic alliances.*
8,249 citations
TL;DR: In this article, the authors examined the links between corporate patent and patent citation data, and several other indicators of corporate performance: changes in sales and profits, research and development budgets, scientific productivity, and expert opinions of company technological strength.
663 citations
TL;DR: In this article, the structural properties of knowledge and collaboration networks and their possible influences on organizational innovations in terms of exploitation and exploration in the emerging nano-energy field were explored. But, the results showed that the knowledge networks and the technology-based collaboration networks are decoupled and that they have different degrees of integration.
320 citations
TL;DR: In this paper, a review of 1660 patents related to biodiesel production were reviewed and grouped into five categories depending on whether they related to starting materials, pre-treatment methods, catalysts, reactors and processing methods or testing methods.
Abstract: Biodiesel is a renewable fuel made from vegetable oils and animal fats. Compared with fossil fuels, it has the potential to alleviate environmental pressures and achieve sustainable development. In this paper, 1660 patents related to biodiesel production were reviewed. They were published between January 1999 and July 2018 and were retrieved from the Derwent Innovation patent database. The patents were grouped into five categories depending on whether they related to starting materials, pre-treatment methods, catalysts, reactors and processing methods, or testing methods. Their analysis shows that the availability of biodiesel starting materials depends on climate, geographical location, local soil conditions, and local agricultural practices. Starting materials constitute 75% of overall production costs and, therefore, it is crucial to select the best feedstock. Pre-treatment of feedstock can improve its suitability for processing and increase extraction effectiveness and oil yield. Catalysts can enhance the solubility of alcohol, leading to higher reaction rates, faster biodiesel production processes, and lower biodiesel production costs. Moreover, the apparatus and processes used strongly affect the oil yield and quality, and production cost. In order to be commercialized and marketed, biodiesel should pass either the American Society for Testing and Materials (ASTM) standards or European Standards (EN). Due to increases in environmental awareness, it is likely that the number of published patents on biodiesel production will remain stable or even increase.
289 citations
TL;DR: In this article, the authors conducted a four-stage Delphi study with 25 experts in order to identify the key drivers, inhibitors and likely future developments in collaborative consumption over the next 10 years.
271 citations