Two critical bottlenecks can be found in the existing expert systems of the rice pest and disease: 1) Low efficiency during knowledge graph (KG) construction can often lead to the entity omission and relationship misjudgment, due to the complex ontologies and overlapping semantics in agricultural texts; 2) In weak generalization in the question-answering (QA), especially for semantically ambiguous queries (e.g., symptom-based disease diagnosis), conventional retrieval-augmented generation (RAG) or natural language-to-Cypher (NL2Cypher) can fail to generate effective queries. In this study, an integrated framework was developed to combine with the Large Language Models (LLMs) and KGs, in order to enhance the fine-grained agricultural knowledge extraction, dynamic KG updates, and precise QA reasoning. Two models were proposed in automated KG construction. 1) Progressive Prompt Extraction (PPE) model was decomposed the entity attribute and triple extraction into three sequential polling stages: corpus polling, which structured unstructured text from the rice pest and disease knowledge base (RDP-KB) into indexable metadata to focus knowledge; entity attribute polling, which used entity attribute ontologies and explanation instructions from an ontology registration module to guide LLM-based attribute extraction; and triple polling, which reused previously extracted entity attributes to enhance relationship extraction. 2) The Incremental Entity Update (IEU) model was addressed on the multi-source data redundancy and conflicts, in order to classify the attributes into "primary attributes" (stable, high-credibility knowledge in the structured database) and "secondary attributes" (newly extracted data from PPE). It prioritized high-reliability sources (e.g., academic literature and professional books over web encyclopedias) and retained primary attributes in case of conflicts. Meanwhile, the AgenticGraphRAG architecture was designed in QA. The ReAct paradigm was adopted to decompose the complex user queries into subtasks. A dual-retrieval system was deployed: vector task retrieval and Cypher task retrieval, thus ensuring complementary and accurate knowledge recall. On the RDP-KB (227 562 characters, constructed via OCR of professional books, web crawling, and manual annotation of 910 entities and 1,602 triples), the framework successfully constructed a KG with 901 entities and 1 554 triples, thus consuming 3 084 874 Tokens and taking 3.38 h. Ablation experiments showed that when Qwen2.5-72B-Instruct was paired with PPE, its F1-scores for entity attribute and triple extraction reached 89.5% and 87.1%, respectively, indicating the increases of 11.6% and 10.5%, compared with the model without PPE. Even the smaller Qwen2.5-32B-Instruct with PPE also outperformed Qwen2.5-72B-Instruct without PPE: Its entity attribute F1-score was 85.1% (vs. 77.9% for Qwen2.5-72B-Instruct w/o PPE), and triple F1-score was 81.4% (vs. 76.6%). Comparative experiments against the baselines (NaiveRAG, GraphRAG, HippoRAG2, LightRAG) revealed that: on the RDP-QA-Symptom dataset (100 real farmer queries + 300 LLM-simulated queries), diagnostic accuracies were 86% (Human subset) and 89.33% (LLM subset), outperforming HippoRAG2 (73%, and 77.33%) and LightRAG (68%, and 70.67%); on the RDP-QA-Web dataset (400 comprehensive QA pairs from agricultural platforms), the system was achieved an LLM-Metric score of 90.32% (vs. 58.16% for NaiveRAG, 65.72% for GraphRAG, 79.56% for HippoRAG2, 82.08% for LightRAG), with notable improvements in accuracy (17.76 vs. 14.96 for LightRAG) and comprehensiveness (18.56 vs. 16.04 for LightRAG). The highest Customer Satisfaction Score (CSAT) of 88% was also achieved among all tests, as evaluated by two human experts. The framework effectively solved the key challenges in the expert systems of the rice pest and disease. The PPE and IEU significantly improved the KG construction efficiency, data consistency, and extraction accuracy, even enabling smaller LLMs to outperform larger ones without PPE. AgenticGraphRAG enhanced QA accuracy in the semantically ambiguous scenarios using dual retrieval. This finding can provide a reusable technical paradigm for the vertical-domain expert systems (e.g., adaptable to other agricultural crops) for future optimizations, such as multi-modal interactions (speech and image). The lower application thresholds can also fully meet the automated KG updates suitable for pest and disease knowledge, further advancing agricultural informatization.

Accurate classification and detection have posed a great challenge on benthic organisms, due to the excessive parameters, computational overhead, and complex underwater environments. In this study, a lightweight target algorithm (named YOLOv7-RFPCW) was proposed to improve the detection accuracy of underwater benthic organisms. 1) The original YOLOv7 network architecture was reconstructed into the feature extraction, in order to significantly reduce the parameters and computational complexity. The overall footprint was effectively shrunk to simultaneously enhance the adaptability in the underwater scenes. This modification served as the lightweight solution. The deployment also remained computationally feasible in the resource-constrained underwater or real-time applications where processing speed was paramount. Progressive Efficient Lightweight Attention Network (P-ELAN) and its variant, P-ELAN-W, were integrated into the architecture to further lighten the network. The existing components were replaced or augmented to streamline the information flow, and then prune unnecessary computations for the essential feature representations. P-ELAN and P-ELAN-W contributed to a substantial reduction in the overall complexity. While the discriminatory power was preserved for the accurate detection of benthic organisms. A convolutional Block Attention Module (CBAM) was employed to recognize the color distortion and spatial localization in the underwater imagery. This attention mechanism was seamlessly integrated into the network to reinforce the feature fusion. The crucial visual cues were discerned in the context of the visually degraded underwater, in order to mitigate the information loss. The attention of CBAM networks was focused on the major features relevant to benthic organisms in the environmental factors of water turbidity and light scattering. 2) The default CIOU loss function was substituted with the Weighted Intersection over Union-Version 3 (WIOU-V3) loss function. The reason was that the shape deformation was commonly encountered with the underwater targets. WIOU-V3 was also tailored to handle the varying shapes and orientations of underwater benthic organisms, in order to provide the adaptive measure of bounding box prediction accuracy. The advanced loss function was incorporated to better cope with the underwater object shapes, thereby reducing the false negatives or positives from the shape-related misinterpretations. Extensive experiment was carried out to evaluate the efficacy of the YOLOv7-RFPCW. The results demonstrate that the parameters were significantly reduced by 75.9%, while there was a substantial decrease of 30.7% in the computational requirements. Moreover, the size of the model was compressed by 75.3%, indicating the profound impact of lightweight strategies. Notably, this efficiency was achieved in the high detection accuracy, which was improved by 1.9 percentage points, indicating successfully balanced compactness with high detection performance. In summary, the YOLOv7-RFPCW algorithm presented as the effective solution to detect underwater benthic organisms. The complexity and computational demands were fully met in the underwater environment, including color distortion, spatial localization errors, and target shape deformations. The improved model can be expected to serve as a robust and efficient tool in the practical deployment for the accurate detection of benthic organisms in underwater scenarios, where the model size and computational overhead were coupled with a tangible increase in detection precision. This finding can offer a reliable means to monitor these vital components of aquatic ecosystems with enhanced efficiency and reliability.