MOTIVATION: Interactions between long noncoding RNAs (lncRNAs) and microRNAs (miRNAs) play pivotal roles in gene regulation and disease progression, notably through mechanisms such as competitive miRNA sponging. Accurate identification of lncRNA-miRNA interactions is therefore essential for understanding disease mechanisms and discovering therapeutic targets. However, current knowledge is largely derived from labor-intensive and costly biological experiments, underscoring the need for reliable computational approaches. RESULTS: We propose DeepLMI, a novel deep learning framework for lncRNA-miRNA interaction prediction that integrates deep feature mining with a globally enhanced graph convolutional network. To effectively capture the distinct properties of lncRNAs and miRNAs, DeepLMI employs specialized feature extraction modules: for lncRNAs, we combine sequence pretraining with self-attention mechanisms to learn multiscale semantic representations; for miRNAs, we fuse heterogeneous features through a graph convolutional encoder. To further address the sparsity and structural complexity of known RNA interaction networks, we design a Global-Enhanced Graph Convolutional Network that jointly models local neighborhood information and global topological signals. The embeddings learned for lncRNAs and miRNAs are then integrated to infer interaction probabilities. Extensive experiments across multiple datasets and evaluation settings demonstrate that DeepLMI consistently outperforms existing state-of-the-art methods and exhibits strong robustness, highlighting its potential as a valuable tool for RNA interaction analysis and disease research. AVAILABILITY AND IMPLEMENTATION: The codes and data are publicly available at https://github.com/Hhhzj-7/DeepLMI.

Automated web GUI testing approaches aim to maximize the code coverage of a web app within a specific time budget. However, due to the highly dynamic characteristics of web apps, testing approaches often get stuck in loops or repeatedly explore the same app areas. To address this issue, existing approaches conduct state abstraction, grouping similar pages into the same state in an effort to approximate the ideal state (i.e., a state that encompasses all-and-only those pages exhibiting the same behavior from a testing perspective) to reduce repetitive explorations. Typically, these approaches rely on the Document Object Model (DOM) or visual similarity, using predefined thresholds or learning-based classifiers to determine which pages should belong to the same state. However, pages within the same ideal state still exhibit discrepancies, caused by factors such as dynamically loaded data and dynamically expanded UI elements. The varying page complexities and design styles among apps bring even more challenges. These phenomena present substantial obstacles to existing approaches in determining desirable classification thresholds or training desirable classifiers, preventing them from conducting satisfactory state abstraction to guide the testing process. To address the preceding challenges, in this article, we propose Judge, a novel approach based on structure merging and contrastive learning for state abstraction. Judge includes a “merge-and-classify” strategy. In the “merge” phase, Judge iterates through the DOM tree of each given page and merges web element siblings that share the same subtree structure into a single one to abstract and simplify the page, while discarding text contents and HTML attributes of web elements in the process. In this way, Judge mitigates the negative effects introduced by dynamically loaded data and dynamically expanded UI elements, substantially reducing discrepancies between pages in the same ideal state. In the “classify” phase, Judge uses a dedicated contrastive learning model to embed simplified page DOMs into vectors and further conducts classification with a Support Vector Machine (SVM), enabling classification in high-dimensional vector space and improving generalizability across diverse web apps. We evaluate Judge against 13 widely used baseline approaches. The results highlight that Judge outperforms these baseline approaches in classifying page pairs, with an average margin ranging from 8.95% to 28.90% in the F1 score across three manually labeled datasets. Additionally, when compared to the five most effective baseline approaches, Judge demonstrates superiority in guiding the exploration of automated web GUI testing in six widely studied apps, with code coverage improved by an average of 2.62–14.12%. The code and data of Judge are publicly accessible.

Environmental, social, and governance (ESG) serves as a crucial indicator for evaluating firms in terms of sustainable development. However, the existing ESG evaluation systems suffer from limitations, such as narrow coverage, subjective bias, and lack of timeliness. Therefore, there is a pressing need to leverage machine learning methods to predict the ESG performance of firms using their publicly available data. Traditional machine learning models encounter the feature imbalance problem due to the heterogeneity in ESG-related features. Common approaches typically involve unfolding all features, thereby granting highdimensional folding features greater exposure and accessibility to downstream models, which results in the neglect of lowdimensional features. To fill the research gap regarding fully using the heterogeneous features of enterprises to enhance AIbased ESG prediction performance, we propose the Feature Balancing Transformer (FeBT), a model based on autoencoders and Transformer blocks. FeBT incorporates a novel feature balancing technique that compresses and enhances high-dimensional features from imbalanced data into low-dimensional representations, thereby ensuring a more balanced impact of high-dimensional and low-dimensional features on the model's performance in the downstream ESG forecasting module. Extensive experiments verified the superior performance of FeBT compared with stateof-the-art methods in real-world ESG-related datasets and evidenced that our feature balancing module provides significant insights from high-dimensional folding features.

Bike Sharing Systems (BSSs) have emerged as an economical and eco-friendly solution to alleviate the last-mile problem in intelligent transport systems. As an exclusive route selection problem in BSSs, Bike Trip Planning (BTP) has a clear goal: to select the available routes with minimum time cost for bike users, while satisfying their basic needs, e.g., the bounded longest walking distance. In this paper, we propose a real-time Lyapunov-based Bike Trip Planning (LBTP) policy that considers the users’ waiting at stations, which has not been sufficiently studied in the literature. Both the total time cost and the service rate of all users are the core criteria of a BSS, so we make use of the Lyapunov optimization theory to make a tradeoff between them. Our policy can achieve self-organizing bike redistribution without extra redistribution budget required. The evaluation results show the superiority of our policy on the both system utility and user service rate compared with the existing BTP policies, and reveal the extra travel time fairness degree among different types of users under our policy.

Although providing exceptional asymptotic performance, trial-and-error based reinforcement learning suffers from being data inefficient, i.e., requiring an excessive amount of trial data for training. In this paper, we propose a data-efficient model-based reinforcement learning algorithm based on the Koopman operator theory. By representing the environment dynamics as a linear dynamical system in a high-dimensional space, the Koopman operator theory allows incorporating effective optimal control methods to produce high-quality trials thus accelerating learning. However, when applying the theory for reinforcement learning, with the sparse and unevenly distributed trial data, it is difficult to learn globally linear representations thus leading to serious model bias. To overcome this problem, we devise a local Koopman operator approach that is tailored for the setup of reinforcement learning. By coupling with deep neural networks and the linear quadratic regulator control, we propose the first Koopman operator model-based reinforcement learning algorithm called deep Koopman reinforcement learning (DKRL). The simulation experiment results show the proposed method can outperform existing approaches and can learn robotics control tasks with a small amount of data.