Data Mining Virtual Contaminated Sites to Develop an Educational Learning Tool supported by a k-NN Machine Learning Predictive Algorithm for use in Environmental Engineering Education
Bibliographic record
Abstract
Recent work by Mumford et al. used high-resolution numerical simulations of contaminated sites to evaluate the state of practice for contaminated site investigation1. These simulations were originally developed in collaboration with academic and industry partners, including the U.S. Department of Defense's Environmental Security Technology Certification Program (ESTCP), as a training tool for environmental monitoring and performance optimization. This project focuses on the development of an algorithm to determine available borehole information in those simulations based on user-input, to help leverage this work to create an educational tool. The algorithm begins by validating user-specified coordinates to ensure they fall within an acceptable range. Leveraging principles from linear algebra and a grid-based mapping system, it identifies the nearest existing borehole on a predefined coordinate grid. To optimize efficiency, the algorithm utilizes a 2-D array to store and retrieve coordinate data. It calculates distances from the input coordinates to all available borehole locations on the grid, to identify the shortest distance and select the closest borehole, enabling students to map contamination within virtual sites. In addition to boreholes, the application incorporates data for Membrane Interface Probes (MIPs), monitoring wells, groundwater samples, and soil samples that are used to investigate contaminated sites, providing a comprehensive tool for training environmental professionals. The k-nearest neighbors (k-NN) algorithm predicts Membrane Interface Probe (MIP) Photoionization Detector (ECD) readings at future locations, which indicate chlorinated contaminant levels. Two models were evaluated: one using spatial data and ECD values, and another incorporating electrical conductivity (EC) and hydraulic conductivity (K), which are also measured by MIP. By making use of additional parameters, variations in soil properties were better represented, leading to more accurate predictions. Through iterative testing and refinement, the tool’s accuracy and user-interface have been enhanced, ensuring robust reliability for classroom applications. These tools are slated for integration into 4th-year subsurface contamination courses at Queen's University, Carleton, and the University of Iowa. By automating borehole selection, instructors can interact more with students and dedicate less time to data processing tasks. The k-NN application will enable the use of machine learning for site assessment to optimize MIP instrument placement to reduce bias. Assistance from Cole Van De Ven (Carleton University) and Jessica Meyer (University of Iowa) in collecting virtual contaminated site data is gratefully acknowledged. References: [1] K. G. Mumford, S. Bryck, B. H. Kueper, S. Mancini, M. Kavanaugh, and D. Reynolds, “Virtual Site Investigation to Evaluate Conceptual Site Model Development at DNAPL ‐Impacted Sites,” Groundwater Monitoring Rem, vol. 42, no. 3, pp. 44–58, Jun. 2022, doi: 10.1111/gwmr.12537.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.003 | 0.003 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.002 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.002 | 0.006 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".