About

John A. Springer, Ph.D., is a Professor in the Department of Computer and Information Technology (CIT) at Purdue University. His main research interests are in Data Science and Artificial Intelligence, with a recent focus on the application of machine and deep learning to Cybersecurity. He serves as the leader of the Purdue DATA Laboratory and holds various leadership roles at Purdue, including co-chair of the CIT Undergraduate Curriculum Committee and co-chair of Purdue’s Integrative Data Science Initiative (IDSI) Curriculum Committee. Dr. Springer has led the national Information Security Research and Education (INSuRE) project from 2016 to 2019, which involves a consortium of universities, governmental organizations, and laboratories engaging students in cybersecurity research problems. His academic career at Purdue includes positions as an Assistant Professor from 2006 to 2011, an Associate Professor from 2011 to 2019, and roles such as Assistant Department Head for Graduate Education and Research in CIT from 2015-2017. He has also served as chair of various curricular and graduate education committees and has been a member of Purdue’s Graduate Council and University Senate. Dr. Springer holds a Ph.D. in Computer Science from Indiana University, along with a Master’s degree in Computer Science and a Bachelor’s degree in Mathematics/Systems Analysis from Taylor University. His industry experience includes roles at Bostech Corporation, Infinite Loop Consulting, Eli Lilly and Company, and Andersen Consulting. He has received multiple awards, including the Purdue University Seed for Success Award, University Faculty Scholar, and Outstanding Faculty awards in his department. His professional affiliations include ACM SIGMOD, ACM, IEEE Computer Society, and IEEE.

Research topics

• Artificial Intelligence
• Computer Science
• Machine Learning
• Aerospace engineering
• Speech recognition
• Distributed computing
• Algorithm
• Telecommunications
• Aeronautics
• Multimedia
• Engineering
• Operations research
• Transport engineering

Selected publications

• Scalable Fuzzy Neural Networks (SFNN) for Multi-Output Human Activity Recognition

  2025-07-06

  article

  Human Activity Recognition (HAR) involves complex, multi-output datasets that demand models balancing efficiency, scalability, and interpretability. This paper proposes a Scalable Fuzzy Neural Network (SFNN), an adaptive hierarchical deep architecture with a multi-section learning mechanism. SFNN processes smaller input windows with effective dimensionality reduction, achieving competitive performance without the heavy computational cost of traditional deep models. Leveraging the transparency of fuzzy systems, it offers a more interpretable alternative to black-box approaches. The model’s theoretical convergence guarantees and strong results on the Opportunity dataset confirm its effectiveness for diverse HAR tasks.

  PublisherDOI

• Detecting Patient and Healthy People’s Personalized Breathing Patterns with Few-Shot Learning

  2025-12-03

  article

  Analysis of respiratory sounds, such as coughing and breathing, has emerged as a promising non-invasive approach for the early detection of pulmonary conditions, including COVID-19 and chronic obstructive pulmonary disease (COPD), as well as for managing treatment plans. In this work, we explore audio-based classification of respiratory conditions in terms of breathing patterns using the few-shot learning approach that requires only a few samples to develop models. We experimented with three publicly available datasets of audio recordings of breathing patterns of healthy people and patients with COVID-19 or COPD. Through a detailed evaluation using three types of audio features commonly employed for audio event classification, with varying embedding dimensions and shot numbers, we found that Mel-Frequency Cepstral Coefficients (MFCCs) with 20 embedding dimensions can achieve an average accuracy of around 85% using only 10 shots when classifying the breathing patterns obtained from the three datasets. These findings highlight the potential for developing audio-based screening tools that require only a few samples, which can be utilized for public health diagnostics.

  PublisherDOI

• Deep Learning in Audio Classification

  Communications in computer and information science · 2022 · 10 citations

  • Computer Science
  • Artificial Intelligence
  • Computer Science
  PublisherDOI

• HAMLET: A Hierarchical Agent-based Machine Learning Platform

  ACM Transactions on Autonomous and Adaptive Systems · 2021 · 5 citations

  • Computer Science
  • Machine Learning
  • Artificial Intelligence

  Hierarchical Multi-agent Systems provide convenient and relevant ways to analyze, model, and simulate complex systems composed of a large number of entities that interact at different levels of abstraction. In this article, we introduce HAMLET (Hierarchical Agent-based Machine LEarning plaTform), a hybrid machine learning platform based on hierarchical multi-agent systems, to facilitate the research and democratization of geographically and/or locally distributed machine learning entities. The proposed system models machine learning solutions as a hypergraph and autonomously sets up a multi-level structure of heterogeneous agents based on their innate capabilities and learned skills. HAMLET aids the design and management of machine learning systems and provides analytical capabilities for research communities to assess the existing and/or new algorithms/datasets through flexible and customizable queries. The proposed hybrid machine learning platform does not assume restrictions on the type of learning algorithms/datasets and is theoretically proven to be sound and complete with polynomial computational requirements. Additionally, it is examined empirically on 120 training and 4 generalized batch testing tasks performed on 24 machine learning algorithms and 9 standard datasets. The provided experimental results not only establish confidence in the platform’s consistency and correctness but also demonstrate its testing and analytical capacity.

  PublisherOA PDFDOI

• Development of a Reliable Method for General Aviation Flight Phase Identification

  IEEE Transactions on Intelligent Transportation Systems · 2021 · 27 citations

  Senior authorCorresponding
  • Computer Science
  • Transport engineering
  • Aeronautics

  Aircraft operations statistics have typically received significant attention from U.S. airport owners and operators and state, local, and federal agencies. Accurate operational data is beneficial in assessing airports’ performance efficiency and impact on the environment, but operational statistics at nontowered general aviation airports are, for the most part, limited or not available. However, the increasing availability and economy of capturing and processing Automatic Dependent Surveillance-Broadcast (ADS-B) data shows promise for improving accessibility to a wide variety of information about the aircraft operating in the vicinity of these airports. Using machine learning technology, specific operational details can be decoded from ADS-B data. This paper aims to develop a reliable and economical method for general aviation aircraft flight phase identification, thereby leading to improved noise and emissions models, which are foundational to addressing many public concerns related to airports.

  PublisherDOI

• Automated Nonlinked Two-Vehicle Towing System

  Communications in computer and information science · 2021-01-01

  book-chapter
  PublisherDOI

• Corrigendum to “Identifying and using driver nodes in temporal networks”

  Journal of Complex Networks · 2019-07-29

  erratumOpen accessSenior author
  PublisherOA PDFDOI

• Identifying and using driver nodes in temporal networks

  Journal of Complex Networks · 2019-01-18 · 19 citations

  articleSenior author

  Abstract In many approaches developed for defining complex networks, the main assumption is that the network is in a relatively stable state that can be approximated with a fixed topology. However, in several applications, this approximation is not adequate because (a) the system modelled is dynamic by nature, and (b) the changes are an essential characteristic that cannot be approximated. Temporal networks capture changes in the topology of networks by including the temporal information associated with their structural connections, that is, links or edges. Here, we focus on controllability of temporal networks, that is, the study of steering the state of a network to any desired state at deadline $t_f$ within $\Delta t=t_f - t_0$ steps through stimulating key nodes called driver nodes. Recent studies provided analytical approaches to find a maximum controllable subspace for an arbitrary set of driver nodes. However, finding the minimum number of driver nodes $N_c$ required to reach full control is computationally prohibitive. In this article, we propose a heuristic algorithm that quickly finds a suboptimal set of driver nodes with size $N_s \geq N_c$. We conduct experiments on synthetic and real-world temporal networks induced from ant colonies and e-mail communications of a manufacturing company. The empirical results in both cases show the heuristic algorithm efficiently identifies a small set of driver nodes that can fully control the networks. Also, as shown in the case of ants’ interactions networks, the driver nodes tend to have a large degree in temporal networks. Furthermore, we analyze the behavior of driver nodes within the context of their datasets, through which, we observe that queen ants tend to avoid becoming a driver node.

  PublisherDOI

• ParBio'17

  2017-08-20

  articleSenior author

  The 6th International Workshop on Parallel and Cloud-based Bioinformatics and Biomedicine (ParBio 2017) aims to bring together scientists in the fields of high performance and cloud computing, computational biology and medicine, to discuss, among the others, the organization of large scale biological and biomedical databases, the parallel and service-based implementation of bioinformatics and biomedical applications, and problems and opportunities of moving biomedical and health applications on the Cloud.

  PublisherDOI

• An impact comparison of two instructional scaffolding strategies employed in our programming laboratories: Employment of a supplemental teaching assistant versus employment of the pair programming methodology

  2017-10-01 · 16 citations

  article

  Instructional scaffolding is a well-researched, commonly-practiced educational technique whereby support is temporarily provided as an individual learns [1]. Grounded in constructivist teaching and learning theory, scaffolding historically referring to support provided by a teacher to a student. However, in the modern learning environment, instructors have access to a wide range of tools and techniques with which to help them scaffold learners [2]. This paper reports the findings of a design experiment comparing the employment of two instructional scaffolding strategies in reducing the impact of large laboratory class size on undergraduate students enrolled in an introductory computer programming course having a mandatory laboratory component. The design experiment compared the performance of students on programming procedural knowledge assessments as well as programming self-beliefs in two offerings of the course, each treated as a cohort, which differed in the instructional scaffolding strategy employed in the mandatory laboratory component. In one cohort, an undergraduate teaching assistant was used as a supplement to the laboratory instructor while students completed programming exercises in the laboratory component of the course. In the other cohort, a cooperative programming methodology known as pair programming was used as a supplement to the laboratory instructor while students completed programming exercises in the laboratory component of the course. Cohort composition, including school of study (e.g., Liberal Arts, Science, Technology, Engineering, Business, etc.), student classification (i.e., freshman, sophomore, junior, senior, or other), gender ratio, and amount of prior programming experience, was comparable. Course content, administrative processes and assessment mechanisms (formative and summative) remained constant. Each student completed a summative assessment of programming procedural knowledge at weeks 7, 12, and 16 of the semester. Open ended feedback was solicited from each student via anonymous questionnaire during weeks 8 and 16 of the semester. Finally, each student completed a 19-item Likert-scale questionnaire investigating their self-beliefs on 5 constructs: debugging self-efficacy, programming self-concept, programming interest, programming anxiety, and programming aptitude mindset. The questionnaire employed to investigate self-beliefs is based on the Scott & Ghinea [3] instrument, modified for use in the specific context of the course. Our results indicate that the two scaffolding strategies provided comparable support to student learning during the first 12 weeks of the semester. However, during the last 4 weeks of the semester, the cohort scaffolded by a supplemental teaching assistant slightly outperformed the cohort scaffolded by employment of the pair programming methodology. This performance difference, though small, was statistically significant. Student programming self-beliefs, however, remained comparable between cohorts throughout the entire semester. Interestingly, the theme of `immediacy of assistance' revealed itself during analysis of student open-ended feedback in both cohorts. These findings are considered using as lens Vygotsky's Zone of Proximal Development theoretical concept [4]. Implications are discussed for instructional practitioners considering employing these scaffoldings strategies in their own learning environments.

  PublisherDOI

Frequent coauthors

• Michael D. Kane

  Rutgers Cancer Institute of New Jersey

  8 shared

• Jon E. Sprague

  Bowling Green State University

  7 shared

• Brandeis Marshall

  3 shared

• Thomas J. Hacker

  Purdue University West Lafayette

  3 shared

• Nicolae Morar

  Deakin University

  3 shared

• Edward L. Robertson

  Suez (Canada)

  3 shared

• Nicholas V. Iannotti

  3 shared

• Kevin Haynes

  Janssen (United States)

  2 shared

Awards & honors

• 2016 - Purdue University Seed for Success Award
• 2014 - University Faculty Scholar
• 2012 - Outstanding Faculty in Discovery Award in Department…
• 2011 - Purdue University Seed for Success Award
• 2011 - Outstanding Non-Tenured Faculty Award in Department o…