Using Physical and Social Sensors in Real-Time Data Streaming for Natural Hazard Monitoring and Response
Bibliographic record
Abstract
Technological breakthroughs in computing over the last few decades have resulted in important advances in natural hazards analysis. In particular, integration of a wide variety of information sources, including observations from spatially-referenced physical sensors and new social media sources, enables better estimates of real-time hazard. The main goal of this work is to utilize innovative streaming algorithms for improved real-time seismic hazard analysis by integrating different data sources and processing tools into cloud applications. In streaming algorithms, a sequence of items from physical and social sensors can be processed in as little as one pass with no need to store the data locally. Massive data volumes can be analyzed in near-real time with reasonable limits on storage space, an important advantage for natural hazard analysis.\nSeismic hazard maps are used by policymakers to set earthquake resistant construction standards, by insurance companies to set insurance rates and by civil engineers to estimate stability and damage potential. This research first focuses on improving probabilistic seismic hazard map production. The result is a series of maps for different frequency bands at significantly increased resolution with much lower latency time that includes a range of high-resolution sensitivity tests.\nSecond, a method is developed for real-time earthquake intensity estimation using joint streaming analysis from physical and social sensors. Automatically calculated intensity estimates from physical sensors such as seismometers use empirical relationships between ground motion and intensity, while those from social sensors employ questionaries that evaluate ground shaking levels based on personal observations. Neither is always sufficiently precise and/or timely. Results demonstrate that joint processing can significantly reduce the response time to a damaging earthquake and estimate preliminary intensity levels during the first ten minutes after an event. The combination of social media and network sensor data, in conjunction with innovative computing algorithms, provides a new paradigm for real-time earthquake detection, facilitating rapid and inexpensive risk reduction. In particular, streaming algorithms are an efficient method that addresses three major problems in hazard estimation by improving resolution, decreasing processing latency to near real-time standards and providing more accurate results through the integration of multiple data sets.
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 machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.003 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 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 source (direct Gemma or distilled Codex), 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".