Our Work

An output of the Unlocking data, https://unlockingdata.africa/ Available from https://docs.unlockingdata.africa/lib/ND47FCTR

Security operations centers must judge every alert on four questions - what tactic is behind it, how urgent, how severe, and how to respond - without letting a single log leave the premises. No existi…

Security operations centers must judge every alert on four questions - what tactic is behind it, how urgent, how severe, and how to respond - without letting a single log leave the premises. No existing approach meets this constraint. Generative LLMs answer uently but are badly calibrated; cloud decision models are trustworthy but demand shipping logs o-site; and open parallel decision models (PDMs) collapse zero-shot on real multi-source logs. The root cause, our audit shows, is not model scale but training data: security-domain post-training on clean raw logs is the missing ingredient. We present ORACLE, a fully local decision layer built on this principle. ORACLE converts four public raw-log corpora into 60k typed-decision questions, post-trains four open PDM lines spanning 0.4B to 4B parameters and three ∗ Corresponding author. Email addresses: xiongxinli_@nudt.edu.cn (Xinli Xiong), huangjun0111@nudt.edu.cn (Jun Huang), wei0427@nudt.edu.cn (Zhihua Wei), wangyongjie17@nudt.edu.cn (Yongjie Wang), niulixian2001@nudt.edu.cn (Lixian Niu) training paradigms, calibrates each question type on target-domain data, and gates every automated action by calibrated con dence, deferring the rest to analysts. On a held-out attack-range test suite, the strongest post-trained line reaches 0.85 tactic accuracy - above the best zero-shot cloud API by a wide margin under an identical protocol - while halving calibration error and keeping the fastest line at 72 ms per decision. ORACLE thus delivers what no prior system could: accurate, calibrated, fast alert adjudication entirely on-premises.

R scripts and machine-readable output tables for the manuscript "Interpreting longitudinal change in the creatinine-to-cystatin C ratio: measurement comparability and model parameterisation in a reana…

R scripts and machine-readable output tables for the manuscript "Interpreting longitudinal change in the creatinine-to-cystatin C ratio: measurement comparability and model parameterisation in a reanalysis of CHARLS". Includes two-wave comparability analyses, uniform-offset and stratum-specific offset scans, and the 2015-landmark reconstruction. See README.md for run order and requirements. Data source: CHARLS public-use data (http://charls.pku.edu.cn).

What's Changed bump by @mrbuche in https://github.com/mrbuche/conspire.rs/pull/224 Exodus reader: accept single-precision coordinates by @mrbuche in https://github.com/mrbuche/conspire.rs/pull/227 F…

What's Changed bump by @mrbuche in https://github.com/mrbuche/conspire.rs/pull/224 Exodus reader: accept single-precision coordinates by @mrbuche in https://github.com/mrbuche/conspire.rs/pull/227 Full Changelog: https://github.com/mrbuche/conspire.rs/compare/v0.7.9...v0.8.0

What's Changed ICOP Subset by @Finn1928 in https://github.com/oral-health-and-disease-ontologies/ohd-ontology/pull/182 Release readme by @Finn1928 in https://github.com/oral-health-and-disease-ontolo…

What's Changed ICOP Subset by @Finn1928 in https://github.com/oral-health-and-disease-ontologies/ohd-ontology/pull/182 Release readme by @Finn1928 in https://github.com/oral-health-and-disease-ontologies/ohd-ontology/pull/183 Icop meta data by @Finn1928 in https://github.com/oral-health-and-disease-ontologies/ohd-ontology/pull/186 insert language tags by @wdduncan in https://github.com/oral-health-and-disease-ontologies/ohd-ontology/pull/192 Issue 190 - add occlusion terms by @wdduncan in https://github.com/oral-health-and-disease-ontologies/ohd-ontology/pull/191 Obo dashboard fixes by @Finn1928 in https://github.com/oral-health-and-disease-ontologies/ohd-ontology/pull/189 fix: some language tags were missed; added them by @wdduncan in https://github.com/oral-health-and-disease-ontologies/ohd-ontology/pull/193 flossing by @Finn1928 in https://github.com/oral-health-and-disease-ontologies/ohd-ontology/pull/195 multiple definitions by @Finn1928 in https://github.com/oral-health-and-disease-ontologies/ohd-ontology/pull/194 Question classes by @Finn1928 in https://github.com/oral-health-and-disease-ontologies/ohd-ontology/pull/188 Emro pain import by @Finn1928 in https://github.com/oral-health-and-disease-ontologies/ohd-ontology/pull/196 Import icop terms from pain by @wdduncan in https://github.com/oral-health-and-disease-ontologies/ohd-ontology/pull/198 removing uneeded files by @wdduncan in https://github.com/oral-health-and-disease-ontologies/ohd-ontology/pull/199 release 2026-10-01 by @wdduncan in https://github.com/oral-health-and-disease-ontologies/ohd-ontology/pull/200 Full Changelog: https://github.com/oral-health-and-disease-ontologies/ohd-ontology/compare/v2026-03-26...v2026-10-01

Introduction: This study aims to evaluate the effectiveness and safety of minimally invasive aesthetic procedures, focusing on the use of PDO threads for facial rejuvenation, comparing them with …

Introduction: This study aims to evaluate the effectiveness and safety of minimally invasive aesthetic procedures, focusing on the use of PDO threads for facial rejuvenation, comparing them with traditional surgical facelift. Methodology: A literature review was conducted to investigate the efficacy of PDO threads, considering aspects such as safety, popularity, and cost-effectiveness compared to surgical techniques. Theoretical Framework: PDO threads are biodegradable monofilaments that stimulate the production of collagen and elastin, providing support and rejuvenation to the skin. They are inserted following the natural tension lines of the skin and have demonstrated efficacy in improving skin laxity and facial rejuvenation. Results and Discussions: The analysis revealed that PDO threads offer a safe and effective alternative for facial rejuvenation, with natural and progressive results. The technique is minimally invasive, with a low complication rate and a faster recovery compared to traditional facelift surgery. Conclusion: Minimally invasive aesthetic procedures with PDO threads are an effective and safe option for facial rejuvenation. They offer significant advantages over surgical facelift, including lower risk of complications and a reduced recovery period. The growing popularity of these procedures reflects the demand for rejuvenation solutions that combine safety, efficacy, and affordability.

Exposición sobre el proceso de reporte de una revisión sistemática. Organización del proceso de identificación y selección de estudios. Elaboración e i…

Exposición sobre el proceso de reporte de una revisión sistemática. Organización del proceso de identificación y selección de estudios. Elaboración e interpretación del diagrama de flujo PRISMA. Revisión de ejemplos de revisiones sistemáticas

Aplicativo web educacional desenvolvido em R e Shiny para realizar operações elementares com matrizes e explorar transformações lineares no plano por meio de uma figura binária.