International Core Journal of Engineering 2020-26 | Page 104
2019 International Conference on Artificial Intelligence and Advanced Manufacturing (AIAM)
Detection of Research Front Topic Based on Data
of NSF Artificial Intelligence Project
Liqu Wen Wei Zhang
Department of domestic Information Research
Beijing Science & Technology Information Research
Institute
Beijing, China
E-mail: wenlq@bjstinfo.com.cn Department of domestic Information Research
Beijing Science & Technology Information Research
Institute
Beijing, China
Xiaohong Jin Ziqiang Liu
Department of domestic Information Research
Beijing Science & Technology Information Research
Institute
Beijing, China Department of Library, Information and Archives
Management, School of Economics and Management
University of Chinese Academy of Sciences
Beijing, China
E-mail: liuziqian@mail.las.ac.cn
Abstract — Artificial intelligence in the United States has
been the world's leading, and the detection of research front
topic based on the data of NSF artificial intelligent projects is
helpful for us to obtain quantitative information for prioritized
technologies that could be used for technology management and
decision making for research funding and technology
investment in the field of artificial intelligent. PLDA model was
used to identify topic contained in the NSF project data in the
field of artificial intelligence from 2013 to 2018. The funding
time, funding amount and centrality indexes of the topic were
taken as three indexes to judge research front topics.
Threshold value was set according to Pareto Law. And 17
research front topics in the artificial intelligence field were
determined. The results showed that 13 aspects, including
intelligent system and intelligent community, quantum
technology, molecular programming, intelligent simulation,
underwater communication and positioning, big data analysis
and so on, are the main research front contents concerned in the
field of artificial intelligence in the United State. co-word analysis, probability theme mode, etc. The effect of
probability theme model is preferably [2].
In 2017, Zhang lei et al [1] used text analysis and tech-
mining methods to analyze the National Science Foundation
(NSF)fund projects data, and illustrates Graphene will impact
the evolution of the main research fields and research hot-
spots. Based on the Top50 subject words obtained, the
emerging indicators were used to explore the technical hot-
spots and future research directions in the Graphene field. In
2017, Wang wenjuan et al. [3] found a new way based on
LDA to detect subjects of projects and their articles to grasp
topic distribution and evolution of funding agency and found
the topics of projects. In 2017, Wang xiaoyue et al. [4] used
PLDA model to identify the theme in the fund project data of
carbon nano-tubes. And they detected the research front
topics based on the subsidy time of topics, the amount of
subsidies and the center of the index. At the same time, the
visual analysis method of topic evolution is used to analyze
the evolution of research topics and predict the development
trend. In 2018, Xu lu et al. [5] proposed a new scientific
research fronts detecting method to fund sponsored projects
based on the TDT (Topic Detection and Tracking) model.
The paper analyzes the funding intensity, time dimension,
topic dimension, and other attributes of fund projects,
constructed the indexes of scientific research front detection
to identify hot-spots front topics, emerging front topics and
future front topics. As the same time, the competitive
situation of scientific research fronts was revealed. In 2018,
Li jing et al. [6] proposed a method of forecasting and
visualizing the emerging topics of fund sponsored projects
based on time series analysis and SVM model. They
constructed the emerging topics detection formula, and use
deep learning algorithms model to predict and analyze the
development trend of new topics. Moreover, visual analysis
software Gephi is applied to visualize and reveal the
competitive trend in Graphene field.
Keywords—National Science Foundation; Fund project;
PLDA; Artificial intelligence; Research front
I. I NTRODUCTION
The change and development of science and technology is
the result of long-term accumulation of basic research. The
National Science Foundation (NSF), founded in 1950, has
been playing an important role in basic scientific research in
United States and even the whole world. Professor Subra
Suresh, President of the NSF, said in his tenure summary in
2013:In more than 60 years, 70% of the researchers who won
the Nobel Prize were related to NSF fund sponsored projects
[1]. That means more than 200 winners whose research have
been funded by NSF. Therefore, investigating the information
contained in the text of the NSF found projects will help us
know well the research front earlier and faster.
Recently, many scholars have carried out fruitful
theoretical exploration and empirical studies about detecting
research front based on found projects information.
Commonly methods used are mainly burst word detection,
978-1-7281-4691-1/19/$31.00 ©2019 IEEE
DOI 10.1109/AIAM48774.2019.00023
In recent years, artificial intelligence has developed
rapidly and achieved remarkable results. Many countries
follow-up laid out the national development strategies in the
82