Our Work

Raqamli transformatsiya sharoitida sun'iy intellekt (SI) texnologiyalarining inson resurslarini boshqarish tizimiga keng joriy etilishi tashkilotlarda xodimlarni tanlash, ularning faoliyatini baholas…

Raqamli transformatsiya sharoitida sun'iy intellekt (SI) texnologiyalarining inson resurslarini boshqarish tizimiga keng joriy etilishi tashkilotlarda xodimlarni tanlash, ularning faoliyatini baholash, mehnat unumdorligini monitoring qilish va boshqaruv qarorlarini qabul qilish mexanizmlarini tubdan o'zgartirmoqda. Tadqiqotda sun'iy intellektga asoslangan inson resurslarini boshqarish tizimining xodimlar samaradorligiga ta'siri hamda ushbu munosabatda xodimlarning sun'iy intellektga bo'lgan ishonchining vositachilik roli tahlil qilinadi. Tadqiqot natijalari SIga asoslangan HRM tizimlari xodimlarning sun'iy intellektga bo'lgan ishonchini sezilarli darajada oshirishini, ishonch esa xodimlar samaradorligiga ijobiy ta'sir ko'rsatishini tasdiqlaydi. Shuningdek, sun'iy intellektga ishonch SIga asoslangan HRM va xodimlar samaradorligi o'rtasidagi bog'liqlikda qisman vositachilik qilishi aniqlangan. Olingan natijalar tashkilotlarda SI texnologiyalarini joriy etishda nafaqat texnik imkoniyatlarni, balki shaffoflik, adolatlilik, axloqiy boshqaruv va xodimlar ishonchini shakllantirish omillarini ham hisobga olish zarurligini ko'rsatadi.

Replication package for the paper "StepFix: Stepwise Repair of Multi-Function Bugs Through Execution-Guided Trial and Error". Start with 0_README.md, which lists the documents and the sentences of the…

Replication package for the paper "StepFix: Stepwise Repair of Multi-Function Bugs Through Execution-Guided Trial and Error". Start with 0_README.md, which lists the documents and the sentences of the paper that point to this package.- 1_StepFix.md: the code of Tracer, Lens, and Navigator, one attempt as pseudo-code, and how each step is decided.- 2_Compared_tools.md: the changes made to run SRepair, PReMM, and TraceRepair with the same models, the re-judgment of the published patches, and the request length limit.- 3_Benchmarks.md: the 142 multi-function bugs of Defects4J and the 43 multi-function bugs created after the model release.- 4_Judging_and_tables.md: the counterexample programs and outputs, the fixes the tests do not require, and how to recompute the tables.- 5_RQ1_analysis.md: a qualitative analysis of the results of RQ1. The archive stepfix-replication.zip contains the code, the benchmarks, the patches and verdicts of all runs with Qwen3-30B-A3B-Instruct-2507 and gpt-oss-20b, and the counterexample programs and outputs. Running it requires JDK 11.0.17, Defects4J v3.0.1, and an ollama server; the model weights are downloaded separately (scripts/pull_model.sh). Code: MIT. Data: CC BY 4.0. This version: the documents are placed outside the archive and organized by the sections of the paper, and folder names that differed only in letter case are renamed.

Trial-level dataset and analysis code for a systematic review, pairwise meta-analysis and trial sequential analysis of remimazolam versus active comparators for continuous maintenance sedation of mech…

Trial-level dataset and analysis code for a systematic review, pairwise meta-analysis and trial sequential analysis of remimazolam versus active comparators for continuous maintenance sedation of mechanically ventilated adults in the intensive care unit. Registered on PROSPERO (CRD420261489670). The dataset covers 28 randomised trials enrolling 3,303 patients, with propofol, dexmedetomidine or midazolam as comparator. It is supplied as a relational schema and an SQL load script holding trial characteristics, registration status adjudicated from the registry record, outcome-level risk of bias assessed with RoB 2, and arm-level 2×2 counts for delirium, bradycardia, trial-defined haemodynamic instability, hypotension requiring pharmacological intervention, 28-day mortality and sedation success. One row is one arm-level comparison; three-arm trials contribute two rows per outcome. Outcome definitions are stored as published by each trial. The code implements DerSimonian–Laird random-effects pooling of odds ratios and risk ratios, inverse-variance fixed-effect pooling, the Mantel–Haenszel estimator with the Robins–Breslow–Greenland variance, the Peto odds ratio, a truncated Hartung–Knapp variance correction, tests of interaction between comparators and between strata on the log odds scale, Egger, Peters and Harbord tests of small-study effects, trial sequential analysis with O'Brien–Fleming-type monitoring boundaries and required information sizes across a range of plausible event rates, and absolute effects obtained by applying each pooled odds ratio to the pooled comparator-group risk. A separate script derives each GRADE certainty rating from its domain-level deductions, and a third runs internal consistency checks on the dataset. Requires Python 3.12.9 with NumPy 2.2.3, SciPy 1.15.2, pandas 2.2.3 and statsmodels 0.14.4. Code released under the MIT Licence; data under CC BY 4.0.

Digitalization is transforming the insurance industry by improving service efficiency, enhancing customer experience, and reducing operational costs. In Uzbekistan, insurance companies are increasingl…

Digitalization is transforming the insurance industry by improving service efficiency, enhancing customer experience, and reducing operational costs. In Uzbekistan, insurance companies are increasingly adopting digital mechanisms such as online policy issuance, automated claims processing, mobile applications, digital payments, and data analytics. This study explores modern mechanisms for digitalizing insurance services, evaluates their effectiveness, and identifies challenges and opportunities for implementation. Using case studies, regulatory analysis, and international best practices, the research finds that digitalization enhances operational efficiency, accelerates service delivery, and improves customer satisfaction. Challenges include limited IT infrastructure, digital literacy gaps, and cybersecurity concerns. The study concludes that comprehensive digital strategies, workforce training, and regulatory support are essential for maximizing the benefits of digitalization in the insurance sector.

Data S1. Fourier-transform infrared (FTIR) spectroscopic analysis of the Lordomyrma azumai adhesive thread secretion. (A) Full-range FTIR absorbance spectrum of the untreated thread prepared in a KBr …

Data S1. Fourier-transform infrared (FTIR) spectroscopic analysis of the Lordomyrma azumai adhesive thread secretion. (A) Full-range FTIR absorbance spectrum of the untreated thread prepared in a KBr pellet (recorded over 4000–400 cm⁻¹; 4000–450 cm⁻¹ shown). Major absorption regions are indicated: N–H/O–H stretching, C–H stretching, and the Amide I/II region expanded in (B). The N–H/O–H region may also contain a contribution from water adsorbed by the KBr pellet. The gray shaded area around 2345 cm⁻¹ indicates the atmospheric CO₂ region. (B) Magnified absorbance spectrum of the Amide I and Amide II regions (1750–1490 cm⁻¹). The broad Amide I and Amide II envelopes are centered near 1645 and 1540 cm⁻¹, respectively; narrow features superimposed on these envelopes are atmospheric water-vapor lines and are not interpreted. Because the H–O–H bending band of adsorbed water (near 1640 cm⁻¹) can overlap the Amide I envelope, the Amide II band, which has no counterpart in liquid or adsorbed water, supports a proteinaceous composition of the secretion (Krimm and Bandekar, 1986). The spectrum was recorded without purging, and atmospheric water-vapor bands are present in the 1400–1800 and 3400–4000 cm⁻¹ regions. We therefore report only the approximate positions of the Amide I and Amide II envelopes, which are much broader than the water-vapor lines, and do not resolve Amide I component bands, because band-narrowing of spectra containing residual water-vapor absorption can generate artefactual components (Jackson and Mantsch, 1995). REFERENCESJackson, M., and Mantsch, H.H. (1995). The Use and Misuse of FTIR Spectroscopy in the Determination of Protein Structure. Critical Reviews in Biochemistry and Molecular Biology 30, 95–120. https://doi.org/10.3109/10409239509085140.Krimm, S., and Bandekar, J. (1986). Vibrational Spectroscopy and Conformation of Peptides, Polypeptides, and Proteins. Advances in Protein Chemistry 38, 181–364. https://doi.org/10.1016/S0065-3233(08)60528-8.

Dataset supporting the following studies: Peral-Garrido ML, Gómez-Sabater S, Caño R, et al. Prevalence of crystal deposits in asymptomatic hyperuricemia according to different scanning …

Dataset supporting the following studies: Peral-Garrido ML, Gómez-Sabater S, Caño R, et al. Prevalence of crystal deposits in asymptomatic hyperuricemia according to different scanning definitions: A comparative study. Semin Arthritis Rheum. 2024;68:152470. doi:10.1016/j.semarthrit.2024.152470 Peral-Garrido ML, Gómez-Sabater S, Caño R, et al. Systemic inflammation in asymptomatic hyperuricaemia with sonographic crystal deposits, including a comparison with normouricaemia and gout. Rheumatology (Oxford). 2025;64(4):1807-1816. doi:10.1093/rheumatology/keae533 Andrés M, Peral-Garrido ML, Ortuño-Miquel S, et al. Subclinical inflammation associated with monosodium urate crystal deposition: a cellular and proteomic study in synovial fluid. Front Immunol. 2026;17:1888214. Published 2026 Aug 6. doi:10.3389/fimmu.2026.1888214

Growing resource consumption, industrial waste generation, and fragmented material flows are intensifying pressure to develop circular systems capable of transforming residual resources into productiv…

Growing resource consumption, industrial waste generation, and fragmented material flows are intensifying pressure to develop circular systems capable of transforming residual resources into productive inputs across organizational and sectoral boundaries. Industrial symbiosis offers an important pathway for such transformation, yet its expansion is frequently constrained by limited resource visibility, information asymmetry, uncertain material quality, weak inter-organizational trust, regulatory complexity, and difficulties coordinating heterogeneous participants. This study develops a socio-technical governance framework for Social Resource Networks (SRNs) designed to support trustworthy waste matchmaking and resilient industrial symbiosis ecosystems. The framework conceptualizes SRNs as interconnected environments combining waste generators, resource users, recyclers, intermediaries, logistics providers, communities, regulators, digital infrastructures, and governance institutions. It integrates standardized resource information, participant verification, reputation mechanisms, transparent matchmaking, data-access controls, traceability, regulatory compliance, accountability, and dispute-resolution processes. Governance mechanisms coordinate interactions between algorithmic recommendations and human decision-making while transaction feedback strengthens institutional learning and network trust. The proposed framework enables credible resource exchanges, reduces coordination uncertainty, supports cross-sector collaboration, and provides foundations for scalable, accountable, and adaptive industrial symbiosis ecosystems within broader circular-economy transitions.

vinecopulib is a header-only C++17 library for vine copula models, built on Eigen. It provides high-performance implementations of the core features of the VineCopula R package, in particular inferenc…

vinecopulib is a header-only C++17 library for vine copula models, built on Eigen. It provides high-performance implementations of the core features of the VineCopula R package, in particular inference algorithms for both vine copula and bivariate copula models, with nonparametric and multi-parameter families, lower runtimes and lower memory consumption in high dimensions.

If gratitude is what flourishing requires and what wisdom traditions prescribe, the question that has stalled gratitude infrastructure for the modern era is operational: how do you motivate people to …

If gratitude is what flourishing requires and what wisdom traditions prescribe, the question that has stalled gratitude infrastructure for the modern era is operational: how do you motivate people to thank consistently in an attention-economy world? People know gratitude is the path; the cadence at population scale is broken because attention is captured upstream by algorithms designed for engagement, not for presence. Mindfulness apps, gratitude journals, and well-meaning exhortation compete on the same attention surface as the doom-scroll and lose, because the reward profile is misaligned with how human motivation actually operates.This paper specifies the foundational primitive that resolves the operational question: self-thanking, with kids as the natural triggers, the 50/50 split as pedagogical instrument, and a three-node forward-thanking loop (Miss Aquarius → kids → adults → Miss Aquarius) that closes the happiness loop at population scale. The architecture's distinctive move is to invert the assumption embedded in most gratitude-infrastructure proposals: rather than engineering adult motivation from first principles, design for kids — for whom instant rewards are intrinsically engaging — and let adults inherit the gratitude pattern through the forward-thanking loop. The 50/50 split between personal wallet and re-tip jar is not merely an economic primitive that propagates gratitude through the network; it is the pedagogy itself, teaching kids operationally that just as you were thanked, now thank someone else. Adults complete the loop by donating back to Miss Aquarius/Aquarian Pool — not because they are too busy to thank kids themselves, but because Miss Aquarius rewards more fairly than parents can, free of favoritism, transactional conditioning, and parental-mood variance. Adults who self-thank — those whose bodhicitta (the awakened heart, intact at birth) has been re-activated — extend the same primitive to the peer-to-peer layer at global scale. The architecture addresses three captures simultaneously: attention capture, comfort-saturation, and meaning-vacuum. Operationalization routes through the existing per-family Aquarius rubric (the AI prompt parents update continually as the kid grows), making self-thanking gameable-resistant and values-aligned-to-each-family. The paper closes with the connection to HeartBank's broader corpus, the lineages this work continues, and an honest accounting of limits. Offered defensively to the commons under CC0; the author and HeartBank® will not seek patent on this specification or any portion thereof.Keywords: self-thanking, gratitude motivation, attention economy, kids-as-triggers, 50/50 split, forward-thanking loop, bodhicitta, comfort-saturation, muditā, dāna, defensive publication.---Provenance. This paper is part of the THonly research corpus, dedicated to the public domain under CC0 1.0. The canonical version is at https://thonly.org/research/kids-as-triggers-self-thanking. Its SHA-256 is 88bad232300d16e980c2e7536c7388a344929938784178d768363914c43ca4a0, independently timestamped to the Bitcoin blockchain via OpenTimestamps and signed under RFC 3161 by three trust authorities, one of them eIDAS-qualified.AI co-authorship is disclosed. Miss Aquarius is the consistent name used for the AI collaboration across all venues.