Search for: All records

The integration of machine learning (ML) and deep learning (DL) into geoscience has experienced a pronounced uptick in recent years, a trend propelled by the intricate nature of geosystems and the abundance of data they produce. These computational methods have been harnessed across a spectrum of geoscientific challenges, from climate modeling to seismic analysis, exhibiting notable efficacy in extracting valuable insights from intricate geological datasets for applications such as mineral prediction. A thorough analysis of the literature indicates a marked escalation in AI-centric geoscience research starting in 2018, characterized by a predictive research orientation and a persistent focus on key computational terms. The thematic network and evolution analyses underscore the enduring prominence of “deep learning” and “machine learning” as pivotal themes, alongside progressive developments in “transfer learning” and “big data”. Despite these advancements, other methodologies have garnered comparatively lesser focus. While ML and DL have registered successes in the realm of mineral prediction, their amalgamation with domain-specific knowledge and symbolic reasoning could further amplify their interpretability and operational efficiency. Neuro-Symbolic AI (NSAI) emerges as a cutting-edge approach that synergizes DL’s robust capabilities with the precision of symbolic reasoning, facilitating the creation of models that are both powerful and interpretable. NSAI distinguishes itself by surmounting traditional ML constraints through the incorporation of expert insights and delivering explanatory power behind its predictive prowess, rendering it particularly advantageous for mineral prediction tasks. This literature review delves into the promising potential of NSAI, alongside ML and DL, within the geoscientific domain, spotlighting mineral prediction as a key area of focus. Despite the hurdles associated with infusing domain expertise into symbolic formats and mitigating biases inherent in symbolic reasoning, the application of NSAI in the realm of critical mineral prediction stands to catalyze a paradigm shift in the field. By bolstering prediction accuracy, enhancing decision-making processes, and fostering sustainable resource exploitation, NSAI holds the potential to significantly reshape geoscience’s future trajectory. 

Abstract Recently artificial intelligence (AI) and machine learning (ML) models have demonstrated remarkable progress with applications developed in various domains. It is also increasingly discussed that AI and ML models and applications should be transparent, explainable, and trustworthy. Accordingly, the field of Explainable AI (XAI) is expanding rapidly. XAI holds substantial promise for improving trust and transparency in AI-based systems by explaining how complex models such as the deep neural network (DNN) produces their outcomes. Moreover, many researchers and practitioners consider that using provenance to explain these complex models will help improve transparency in AI-based systems. In this paper, we conduct a systematic literature review of provenance, XAI, and trustworthy AI (TAI) to explain the fundamental concepts and illustrate the potential of using provenance as a medium to help accomplish explainability in AI-based systems. Moreover, we also discuss the patterns of recent developments in this area and offer a vision for research in the near future. We hope this literature review will serve as a starting point for scholars and practitioners interested in learning about essential components of provenance, XAI, and TAI. 

Abstract Evaluating metagenomic software is key for optimizing metagenome interpretation and focus of the Initiative for the Critical Assessment of Metagenome Interpretation (CAMI). The CAMI II challenge engaged the community to assess methods on realistic and complex datasets with long- and short-read sequences, created computationally from around 1,700 new and known genomes, as well as 600 new plasmids and viruses. Here we analyze 5,002 results by 76 program versions. Substantial improvements were seen in assembly, some due to long-read data. Related strains still were challenging for assembly and genome recovery through binning, as was assembly quality for the latter. Profilers markedly matured, with taxon profilers and binners excelling at higher bacterial ranks, but underperforming for viruses and Archaea. Clinical pathogen detection results revealed a need to improve reproducibility. Runtime and memory usage analyses identified efficient programs, including top performers with other metrics. The results identify challenges and guide researchers in selecting methods for analyses.