Current crop disease VQA models primarily focus on object counting and detection. However, accurately identifying various disease stages and determining control measures re quire additional knowledge beyond images, including information about control methods and pathogen details. To address this, the VQA dataset relies on the images and questions to retrieve relevant external knowledge. To realize the VQA task of crop diseases external knowledge, we construct the Visual Question Answer Model Based on Crop Diseases External Knowledge for Smart Agriculture (CDEK). CDEK integrates two categories of external knowledge on 66 common dicotyledonous crop diseases by utilizing large language models and agricultural knowledge repositories to enhance knowledge retrieval. This integration enhances the richness of external knowledge repositories. En hancing fine-grained image understanding in CDEK through the utilization of Stack Self-Attention (SSA), utilising Cross Attention and contrastive learning of two external knowledge, with a focus on emphasizing image-related semantic information during training. Finally, an automatic patrol disease detection robot is constructed based on Tensor Processing Unit (TPU) devices and the CDEK model. CDEK achieves an accuracy of 61.7% on the publicly available dataset OK-VQA, surpassing the previous state-of-the-art by 5.1%. Furthermore, we construct the OKiCD-VQA dataset for crop diseases external knowledge and achieve an accuracy of 89.36% using CDEK. A series of ablation experiments are conducted on various modules, the effectiveness of CDEK is demonstrated through extensive experimentation. Contributing solutions to the sustainable development of smart agriculture.

Manufacturing process discovery extracts insights from event logs recorded by Manufacturing Information Systems (MISs) to optimize operational processes. However existing process discovery techniques struggle with complex concurrency relations, resulting in imprecise sub-processes that compromise model accuracy. This paper proposes a novel enhancement to Inductive Miner (IM)-generated models by optimizing local imprecise structures in manufacturing process models. The method first identifies imprecise sub-processes and extracts their corresponding sub-logs. Then imprecise sub-processes are incrementally optimized using a frequency-based filter mechanism, generating multiple candidate models. Finally, the best-quality candidate model based on evaluation metrics is selected as the final output. The proposed technique has been implemented as an open source process mining toolkit ProM plugin and evaluated on six real-life manufacturing event logs. Experimental results demonstrate that it outperforms state-of-the-art techniques, producing higher quality process models, making it particularly suited for manufacturing process discovery.

This paper establishes an analytical model for a dual-pressure-actuated pneumatic linear actuator, investigating the relationship between the output force of the linear actuator and both the pressure differential and displacement. Experiments were designed to validate the model. The maximum output force of the linear actuator under negative pressure (-40 kPa) is 100 N, while under hybrid air pressure (negative pressure -40kPa combined with positive pressure 40 kPa), the maximum output force significantly increases to approximately 210 N, demonstrating that dual pressure driving can substantially enhance output performance. The analytical results exhibit excellent agreement with experimental data under low-pressure conditions, with a maximum relative error of only 5%. Furthermore, comparisons with a flexible bellows of the same dimensions confirm that the linear actuator also exhibits high stiffness. Finally, potential applications of the linear actuator in daily life are discussed.

ABSTRACT Pedestrian detection is crucial in agricultural environments to ensure the safe operation of intelligent machinery. In orchards, pedestrians exhibit unpredictable behavior and can pose significant challenges to navigation and operation. This demands reliable detection technologies that ensures safety while addressing the unique challenges of orchard environments, such as dense foliage, uneven terrain, and varying lighting conditions. To address this, we propose ReB‐DINO, a robust and accurate orchard pedestrian detection model based on an improved DINO. Initially, we improve the feature extraction module of DINO using structural re‐parameterization, enhancing accuracy and speed of the model during training and inference decoupling. In addition, a progressive feature fusion module is employed to fuse the extracted features and improve model accuracy. Finally, the network incorporates a convolutional block attention mechanism and an improved loss function to improve pedestrian detection rates. The experimental results demonstrate a 1.6% improvement in Recall on the NREC dataset compared to the baseline. Moreover, the results show a 4.2% improvement in and the number of parameters decreases by 40.2% compared to the original DINO. In the PiFO dataset, the with a threshold of 0.5 reaches 99.4%, demonstrating high detection accuracy in realistic scenarios. Therefore, our model enhances both detection accuracy and real‐time object detection capabilities in apple orchards, maintaining a lightweight attributes, surpassing mainstream object detection models.

Request for Comment (RFC) is a universal and standardized specification and describes internet protocol processes, algorithms, and standards. The formal representations of network protocols extracted from RFC can be used to verify correctness and security of the communication process. This paper proposes automatically extracting Petri net models from RFC protocol texts, transforming unstructured protocol text into a structured model. First, the RFC protocol text is preprocessed with chunking, part-of-speech, and element labeling. Second, element recognition is performed by the trained Bert-DGCNN-Bi-LSTM-CRF network, and a structured intermediate representation is generated based on the recognition results and semantic role relationships. Then, state transfer relations are extracted from the intermediate representation and stored in the correlation matrix. The correlation matrix is converted to a PNML file for visualization of Petri nets using PIPE software. Finally, an experimental comparative analysis of element recognition and relation extraction is conducted to prove the effectiveness of the extraction approach. Comparing the similarity between automatic and manual model extraction proves the quality of the extracted models.

Existing pest image classification models require a large number of labeled training images. However, labels for most pest images in the real world do not exist. Therefore, the zero-shot learning method based on generative meta-learning provides an effective solution, which first uses attributes to transfer knowledge from seen classes to unseen classes, and then synthesizes the features of unseen classes. We observe that seen and unseen classes share the same high-level attributes, which can be used to learn a shared set of optimal parameters for seen and unseen classes. Therefore, we propose a novel Hierarchical Class level Attribute guided Generative meta model for pest image Zero-shot Learning (HCAG-ZSL). HCAG-ZSL uses the pre-built Taxonomic Attribute Tree to get the high-level attributes corresponding to the class attributes. These attributes are then fed into a well-designed generator to generate visual features. Extensive experiments show that the proposed model outperforms state-of-the-art generative meta models.

As an important task in business process management, remaining time prediction for business process instances has attracted extensive attentions. However, most of the traditional remaining time prediction approaches only take into account formal process models and cannot handle large-scale event logs in an effective manner. Although machine learning and deep learning have been recently applied to the remaining time prediction task, these approaches cannot incorporate domain knowledge naturally. To overcome these weaknesses of existing studies, we propose a remaining execution time prediction approach based on a novel auto-encoded transition system, which can enhance the complementarity of process modeling and deep learning techniques. Through auto-encoding the event-level and state-level features, the proposed approach can represent process instances in a comprehensive and compact form. Furthermore, a transfer learning strategy is proposed to train the remaining time prediction model so as to avoid overfitting and improve the accuracy of prediction. We conduct extensive experiments on four real-world datasets to verify the effectiveness of the proposed approach. The results show its superiority over several state-of-the-art approaches.

Domain terminologies are a basic resource for various natural language processing tasks. To automatically discover terminologies for a domain of interest, most traditional approaches mostly rely on a domain-specific corpus given in advance; thus, the performance of traditional approaches can only be guaranteed when collecting a high-quality domain-specific corpus, which requires extensive human involvement and domain expertise. In this article, we propose a novel approach that is capable of automatically mining domain terminologies using search engine's query log—a type of domain-independent corpus of higher availability, coverage, and timeliness than a manually collected domain-specific corpus. In particular, we represent query log as a heterogeneous network and formulate the task of mining domain terminology as transductive learning on the heterogeneous network. In the proposed approach, the manifold structure of domain-specificity inherent in query log is captured by using a novel network embedding algorithm and further exploited to reduce the need for the manual annotation efforts for domain terminology classification. We select Agriculture and Healthcare as the target domains and experiment using a real query log from a commercial search engine. Experimental results show that the proposed approach outperforms several state-of-the-art approaches.

This paper points out that many machine learning problems in IR should be and can be formalized in a novel way, referred to as 'group-based learning'. In group-based learning, it is assumed that training data as well as testing data consist of groups. The classifier is created and utilized across groups. Furthermore, evaluation in testing and also in training are conducted at group level, with the use of evaluation measures defined on a group. This paper addresses the problem and presents a Boosting algorithm to perform the new learning task. The algorithm, referred to as AdaBoost.Group, is proved to be able to improve accuracies in terms of group-based measures during training.

This paper proposes a new ranking approach for information retrieval, where the diversity among queries are taken into consideration. In information retrieval, the users' queries often vary a lot from one to another, so that the documents retrieved from different queries are also distributed differently. Due to this diversity, it is not appropriate to assume all the documents to be ranked are generated i.i.d. (independently and identically distributed) according to a fixed but unknown probability distribution. However, most of the existing learning to rank approaches are proposed on the basis of the conventional i.i.d. assumption. In this paper, the conventional i.i.d. assumption is relaxed to fit the real situations of information retrieval better, and then a new ranking approach, referred to as 'query dependent ranking', is proposed. In our approach, the ranking models for different queries have generality while each of them has its own speciality. The experimental results on both synthetic and real-world datasets show the advantage of our approach to conventional ranking approaches.