Add BrightSurf on Google Email

Higher Education Press


New framework shows brute-force search is inevitable for certain Boolean puzzles, improving cryptography and AI strategies

Researchers demonstrate that brute-force search is necessary for specific computer puzzles, highlighting the importance of focusing on tractable instance classes. The study's findings have implications for fields like encryption, artificial intelligence, and scheduling.

SourceHigher Education Press·JournalFrontiers of Computer Science·TypeExperimental study·DateJul 3, 2025

Programmable all-optical signal processing on a silicon chip

A research team has developed a monolithically integrated programmable all-optical signal processing chip with filtering, regeneration, and logic operation functions. The chip harnesses the advantages of silicon photonics to deliver high-speed performance, advanced modulation formats, and wavelength transparency. The technology paves t...

SourceHigher Education Press·JournalFrontiers of Optoelectronics·TypeExperimental study·DateJul 2, 2025

AI speeds up accurate prediction of molecular properties for drug discovery

A new AI framework called MetaGIN delivers fast and accurate predictions of molecular properties without relying on complex 3D structural data. This innovation promises to accelerate early-stage drug discovery and bring new therapies closer to patients by reducing computational bottlenecks and resource requirements.

SourceHigher Education Press·JournalFrontiers of Computer Science·TypeExperimental study·DateJul 1, 2025

New method filters noisy data for safer AI application

Researchers at Huazhong University of Science and Technology have developed a new approach to help artificial intelligence work more reliably with messy or misleading data. The Soft-GNN method selectively uses cleaner data during training, achieving lower error rates and more stable performance, particularly under high-noise conditions.

SourceHigher Education Press·JournalFrontiers of Computer Science·TypeExperimental study·DateJul 1, 2025

Clustering-based approach accelerates AI learning in robotics and gaming

Researchers developed Clustered Reinforcement Learning (CRL), a plug-and-play framework that sorts similar situations into clusters, rewarding AI for trying new things and building on past successes. This approach achieves top performance across multiple benchmarks, including robotic control tasks and difficult Atari games.

SourceHigher Education Press·JournalFrontiers of Computer Science·TypeExperimental study·DateJul 1, 2025

New model extracts sentence-level proof to verify events, boosting fact-checking accuracy for journalists, legal teams, and policymakers

A new neural network improves overall fact-checking accuracy and exact-match accuracy on standard benchmarks, highlighting the precise text that supports an event's truth. The model achieves a factual-accuracy score of 66.9% and exact matches rose to 42.9%, demonstrating its ability to adapt across languages.

SourceHigher Education Press·JournalFrontiers of Computer Science·TypeExperimental study·DateJul 1, 2025

Smart algorithms that prevent digital traffic jams—especially when public emergencies strike

Researchers developed an energy-efficient strategy to tackle digital traffic jams during peak hours, ensuring seamless services like navigation updates and pollution alerts. The approach, powered by Lyapunov optimization and the Kuhn–Munkres algorithm, reduces processing delays and power use while maintaining system reliability.

SourceHigher Education Press·JournalFrontiers of Computer Science·TypeExperimental study·DateJul 1, 2025

Predicting miRNA-drug interactions via dual-channel network based on TCN and BiLSTM

A new research published in Frontiers of Computer Science proposes a dual-channel network model to predict miRNA-drug interactions. The model achieves high accuracy using Temporal Convolutional Network and BiLSTM algorithms, with average AUC scores of 0.9567, 0.9365, and 0.8975 on three different datasets.

SourceHigher Education Press·JournalFrontiers of Computer Science·TypeExperimental study·DateJun 23, 2025

Interpretation on feature groups for tree models

A new research method effectively interprets feature groups in tree models by leveraging inherent correlations and structures among multiple features. The approach measures the importance of feature groups using a novel metric called BGShapvalue, which is then used to identify salient feature groups with large values.

SourceHigher Education Press·JournalFrontiers of Computer Science·TypeExperimental study·DateJun 23, 2025

Formamide-engineered VOPO4 cathodes with high volumetric capacity and mass loading for aqueous zinc-ion batteries

This study introduces formamide (FA)-intercalated VOPO₄ nanosheets, which enhance structural stability and ion transport pathways. The resulting electrodes deliver remarkable electrochemical performance, including a specific mass capacity of 463 mAh/g and a volumetric capacity of 733 mAh/cm³.

SourceHigher Education Press·JournalFrontiers in Energy·TypeExperimental study·DateJun 23, 2025

System log isolation for containers

Researchers propose a solution to system log isolation problems in containers by introducing private logs (POGs), which provide individual log configuration, storage, and view for each container. This approach addresses configuration conflict, operation conflict, namespace escape, information leakage, and log redundancy.

SourceHigher Education Press·JournalFrontiers of Computer Science·TypeExperimental study·DateJun 23, 2025

How will the “water footprint” of Xinjiang cotton change under climate change?

A study simulates three irrigation technologies in Xinjiang's driest region, predicting a decrease in the total water footprint of cotton production. Sprinkler irrigation demonstrates the most notable water-saving effects, while furrow and micro-irrigation show relatively smaller reductions.

SourceHigher Education Press·JournalFrontiers of Agricultural Science and Engineering·TypeExperimental study·DateJun 23, 2025

Which legume crop rotation pattern better promotes soil health?

Researchers analyzed 261 Chinese soil samples to compare legume crop rotations, revealing faba bean rotation's benefits in enhancing soil properties and microbial communities. Faba bean rotation increased organic carbon, total nitrogen, and phosphorus, as well as microbial biomass and respiration rate.

SourceHigher Education Press·JournalFrontiers of Agricultural Science and Engineering·TypeExperimental study·DateJun 23, 2025

How has agricultural carbon emissions in Fujian evolved over 20 years?

Agricultural carbon emissions in Fujian exhibited a 'double decline' trend from 2002 to 2022, with total emissions fluctuating downward by 11% and carbon emission intensity plummeting by 82.46%. Promoting organic fertilizers, optimizing rice planting patterns, and improving resource utilization can further reduce emissions.

SourceHigher Education Press·JournalFrontiers of Agricultural Science and Engineering·TypeExperimental study·DateJun 23, 2025

Microbial carbon fixation technology helps CO2 emission reduction in cement industry, and high-value utilization of steel slag ushers in innovative solutions

A recent study employs microbial technology combined with a rotary kiln process to accelerate CO2 fixation from cement kiln flue gas, fixing over 10% of CO2 within an hour. The approach results in improved soundness and hydration activity of steel slag, enabling its safe utilization in construction materials sector.

SourceHigher Education Press·JournalEngineering·DateJun 20, 2025

Intelligent fault diagnosis for CNC through the integration of Large Language Models and domain knowledge graphs

Researchers propose a retrieval-augmented generation method based on LLMs and domain KG, achieving surpassing diagnostic capabilities of experienced engineers. The system has been integrated into the CNC Cloud Manager APP, addressing challenges in symbolic reasoning and providing a standardized framework for industrial applications.

SourceHigher Education Press·JournalEngineering·DateJun 20, 2025

How to achieve intelligent and precise pesticide application in sustainable agriculture?

Researchers develop innovative technology to control black rot disease in cauliflower, achieving high precision and reducing pesticide use by 72.5%. The system uses spectral sensors, machine learning models, and intelligent spraying, successfully identifying and treating diseased plants while avoiding mis-spraying of healthy ones.

SourceHigher Education Press·JournalFrontiers of Agricultural Science and Engineering·TypeExperimental study·DateJun 18, 2025

Water erosion models in China: how to overcome current limitations under an empirical dominance?

Water erosion models in China have been dominated by empirical approaches, but these lack reliable validation and are limited to specific regions. Research has identified challenges, including a verification gap and regional limitation, which hinder accurate predictions and optimization.

SourceHigher Education Press·JournalFrontiers of Agricultural Science and Engineering·TypeExperimental study·DateJun 18, 2025