Donghua Zhu’s research while affiliated with Beijing Institute of Technology and other places

Understanding the evolution of knowledge has been and will continue to be the key task of science, technology, and innovation management. Existing research on evolutionary path identification relies primarily on traditional co-occurrence analysis and bag-of-words (BOW)–based models for topic extraction. However, these approaches have limitations in effectively capturing the underlying semantics and linkages of the topics. In this article, we propose a novel embedding-based methodology for scientific evolution analysis, in which word embedding, document embedding, clustering, and network analysis are applied to extract topics, measure topical semantic similarities, and quantitatively distinguish topics’ evolutionary states. We first perform benchmark experiments to demonstrate that doc2vec generally outperforms the BOW-based models in topic extraction before evolution analysis. We then consider topic consistency in vector spaces to identify evolutionary states including newborn, convergence, inheritance, and extinction. Scientific evolutionary paths are finally unraveled based on topic similarity matrixes and evolutionary states. We conduct a case study on object detection research to validate the effectiveness of our methodology. The empirical results, validated by domain experts, demonstrate that the proposed methodology is capable of effectively revealing patterns of knowledge inheritance and integration. Consequently, this methodology can be used to improve decision-making processes in future innovation management.

Peer review is one of the important procedures to determine which research proposals are to be funded and to evaluate the quality of scientific research. How to find suitable reviewers for scientific research proposals is an important task for funding agencies. Traditional methods for reviewer recommendation focus on the relevance of the proposal and knowledge of candidate reviewers by mainly matching the keywords or disciplines. However, the sparsity of keyword space and the broadness of disciplines lead to inaccurate reviewer recommendations. To overcome these limitations, this paper introduces a reviewer recommendation method (RRM) for scientific research proposals. This research applies word embedding to construct vector representation for terms, which provides a semantic and syntactic measurement. Further, we develop representation models for reviewers’ knowledge and proposals, and recommend reviewers by matching two representation models incorporating ranking fusions. The proposed method is implemented and tested by recommending reviewers for scientific research proposals of the National Natural Science Foundation of China. This research invites reviewers to provide feedback, which works as the benchmark for evaluation. We construct three evaluation metrics, Precision, Strict-precision, and Recall. The results show that the proposed reviewer recommendation method highly improves the accuracy. Research results can provide feasible options for the decision-making of the committee, and improve the efficiency of funding agencies.

Technology opportunity analysis (TOA) is of great help to technology innovation and R&D strategy of enterprises. Most previous studies of TOA only focused on scientific research or technology development phases, seldom linking technology opportunities to business applications. Using both patent and trademark data, we focused on the linkage of technology and business areas and discovered technology opportunities based on firms’ existing technology base. First, we extracted common product keywords from two data sources to construct document-keyword matrices. Then, we matched patents and trademarks with common keywords to build the relevancy network between technology and the business areas. Finally, we discovered technology opportunities from potential undeveloped business areas, taking into account technology-business relevancy and firms’ technology base. We conducted a case study on 3D printing to test our method.

Citation counts are commonly used to evaluate the scientific impact of a publication on the general premise that more citations probably mean more endorsements. However, two questionable assumptions underpin this idea: a) that all authors contributed equally to the paper; and b) that the endorsement is positive. Obviously, neither of these assumptions hold true. Hence, with this study, we examine two components of citations—their purpose, i.e., the reason for the citation, and polarity, being the author’s attitude toward the cited work. Our findings provide a new perspective on the scientific impact of highly-cited publications. Our methodology consists of three steps. Firstly, a pre-trained model composed of a Word2Vec—a well-known word embedding approach—and a convolutional neural network (CNN) is used to identify citation polarity and purpose. Secondly, in a set of highly-cited papers, we compare eight categories of purpose from foundational to critical and three categories of polarity: positive, negative, and neutral. We further explore how different types of papers—those discussing discoveries or those discussing utilitarian topics—influence the evaluation of scientific impact of papers. Finally, we mine and discover the knowledge (e.g. method, concept, tool or data) to explain the actual scientific impact of a highly-cited paper. To demonstrate how combining citation polarity with purpose can provide far greater details of a paper’s scientific impact, we undertake a case study with 370 highly-cited journal articles spanning “Biochemistry & Molecular Biology” and “Genetics & Heredity”. The results yield valuable insights into the assumption about citation counts as a metric for evaluating scientific impact.

To explore the pathogenic mechanisms of MicroRNA (miRNA) on diverse diseases, many researchers have concentrated on discovering the potential associations between miRNA and disease using machine learning methods. However, the prediction accuracy of supervised machine learning methods is limited by lacking of experimentally-validated uncorrelated miRNA-disease pairs. Without these negative samples, training a highly accurate model is much more difficult. Different from traditional miRNA-disease prediction models using randomly selected unknown samples as negative training samples, we propose an ensemble learning framework to solve this positive-unlabeled (PU) learning problem. The framework incorporates two steps, i.e., a novel semi-supervised Kmeans (SS-Kmeans) to extract reliable negative samples from unknown miRNA-disease pairs and subagging method to generate diverse training sample sets to make full use of those reliable negative samples for ensemble learning. Combined with effective random vector functional link (RVFL) network as prediction model, the proposed framework showed superior prediction accuracy comparing with other popular approaches. A case study on lung and gastric neoplasms further confirms the framework’s efficacy at identifying miRNA disease associations.

Formulating good R&D strategy requires sound knowledge of the past and present R&D trends in various industry sectors. Therefore, this paper outlines a framework for mining industry level R&D trends from patents that were designed for enterprises. Unlike the current alternatives, the approach presented here considers both patent applications and invalidated patents, i.e., those patents that have expired, lapsed, or been revoked. The result is a richer and more comprehensive analysis that covers the full lifespan of a targeted technology from emergence to decline. The framework comprises of a LDA topic model that identifies the technologies and sub-technologies, and of each individual patent and invalidated patent. Then, two specifically designed measures chart the stages of the technologies’ life. An application metric reflects annual levels of interest in an area, while an invalidation metric traces waning interest. The output is a series of trend maps that show the levels of interest and disinterest in different avenues of inquiry over time. Charted on different axes, these two metrics create two distinct trend lines that reflect the different changes over a technology's lifecycle. A case study that focused on China's 3-D printing technology illustrates the approach. The analysis results are highly consistent with the present technology trends across industries, which indicates that the method could serve as a useful reference tool for analyzing R&D trends and creating new R&D strategies.

Technology is a key factor for the competitiveness of the world-leading companies. Collaborating with and funding other research organisations are effective and economical ways to improve their innovation performance. This study aims to investigate the roles of companies in scientific research, as an author and as a funder, and the inner links. Big companies in the agrobiotechnology sector have made great efforts to drive the development of genetically modified techniques. Genetically modified maize, as one of the commercialised genetically modified organisms, is selected as the case study. From an investigation of data available from the Web of Science database, we find that companies’ publications are important in terms of quality and quantity. Companies collaborate with various types of organisations, and build international funding networks. These publications involve in a wide range of emerging topics, and company-authored publications have higher citations. Implications for further research are discussed.

Technological innovation is a dynamic process that spans the life cycle of an idea, from scientific research to production. Within this process, there are often a few key innovations that significantly impact a technology's development, and the ability to identify and trace the development of these key innovations comes with a great payoff for researchers and technology managers. In this article, we present a framework for identifying the technology's main evolutionary pathway. What is unique about this framework is that we introduce new indicators that reflect the connectivity and the modularity in the interior citation network to distinguish between the stages of a technology's development. We also show how information about a family of patents can be used to build a comprehensive patent citation network. Finally, we apply integrated approaches of main path analysis (MPA)—namely global MPA and global key-route main analysis—for extracting technological trajectories at different technological stages. We illustrate this approach with dye-sensitized solar cells (DSSCs), a low-cost solar cell belonging to the group of thin-film solar cells, contributing to the remarkable growth in the renewable energy industry. The results show how this approach can trace the main development trajectory of a research field and distinguish key technologies to help decision makers manage the technological stages of their innovation processes more effectively.