Generative List Recommendation (GLR) models have shown superior performance in E-commerce by directly generating high-quality recommendation lists through sequential item selection. While Online Learning (OL) has proven valuable for streaming point-wise recommendation models in adapting to dynamic user preferences, its application to GLR remains largely unexplored, due in large part to the inefficiency and instability of conventional on-policy reinforcement learning algorithms used in existing GLR approaches. Existing approaches typically rely on surrogate losses, which provide indirect and biased gradient estimates, making them ill-suited for the rapid, subtle distribution shifts common in real-world E-commerce environments. In this paper, we propose Environment Policy Optimization (EPO), a novel GLR model that fundamentally reshapes policy learning by exploiting the differentiability of the environment within the Generator-Evaluator framework. EPO recognizes that the evaluator is a neural network, capable of providing gradient signals. By directly utilizing these gradients, EPO enables end-to-end optimization of the total list-wise reward—the true objective. To ensure differentiable list generation, EPO introduces a new indexing and exploration strategy based on NeuralSort and Gumbel noise, which relaxes discrete item selection into a continuous, gradient-friendly operation. EPO not only demonstrates strong performance in offline evaluations but also unlocks the potential of online learning for GLR at an industrial scale, yielding a 1.18% relative improvement in user clicks in online A/B tests. The results underscore the critical role of EPO in providing the sensitivity, stability, and gradient fidelity necessary for effective real-time adaptation in streaming and dynamic recommendation environments. Moreover, EPO reached the baseline performance with a training time deduction of 76% under identical hardware conditions.

To overcome the limitations of traditional satellite navigation technologies in complex and signal-obstructed industrial environments, this paper presents MSM-Pnet, a novel semi-supervised FM-based positioning framework leveraging FM signals of opportunity. By integrating wavelet packet decomposition with a multi-scale Vision Transformer and a hybrid masking strategy that combines random and time–frequency-aware masking, MSM-Pnet introduces a masked autoencoder architecture capable of robust positioning with limited labeled data. Experimental results demonstrate that MSM-Pnet consistently outperforms conventional supervised learning methods in both indoor and outdoor environments, while also significantly reducing model complexity. These results highlight the method’s potential as a cost-effective and scalable solution for seamless indoor–outdoor positioning for Internet of Things systems.

Question-answering (QA) is a natural approach for humans to understand a piece of music audio. However, for machines, accessing a large-scale dataset covering diverse aspects of music is crucial, yet challenging, due to the scarcity of publicly available music data of this type. This paper introduces MQAD, a music QA dataset built on the Million Song Dataset (MSD), encompassing a rich array of musical features - including beat, chord, key, structure, instrument, and genre — across 270,000 tracks, featuring nearly 3 million diverse questions and captions. MQAD distinguishes itself by offering detailed time-varying musical information such as chords and sections, enabling exploration into the inherent structure of music within a song. To compile MQAD, our methodology leverages specialized Music Information Retrieval (MIR) models to extract higher-level musical features and Large Language Models (LLMs) to generate natural language QA pairs. Then, we leverage a multimodal LLM that integrates the LLaMA2 and Whisper architectures, along with novel subjective metrics to assess the performance of MQAD. In experiments, our model trained on MQAD demonstrates advancements over conventional music audio captioning approaches. The dataset and codes are at https://github.com/oyzh888/MQAD.

Accurate and efficient positioning in complex environments remains a critical challenge where satellite-based systems (e.g., GNSS) suffer from signal attenuation and multipath interference. This paper proposes WK-Pnet, a lightweight positioning framework that utilizes frequency modulation (FM) signals and integrates Wavelet Packet Decomposition (WPD) with knowledge distillation. WK-Pnet first decomposes raw FM IQ signals using WPD to extract fine-grained multi-scale time-frequency features, preserving both spectral and phase information. These features are then fed into a deep neural network for location estimation. To reduce computational complexity, we employ a knowledge distillation strategy that transfers knowledge from a large-capacity ResNeXt-based teacher model—enhanced with a spatial attention mechanism—to a compact student network with significantly fewer parameters and FLOPs. The proposed method is validated on publicly available indoor and outdoor datasets, showing that WK-Pnet achieves comparable positioning accuracy to the teacher model while reducing FLOPs by 95.9%, model parameters by 99.3%, and inference latency by 90.5% on edge devices. Experimental comparisons also reveal that WPD outperforms STFT and EMD in positioning stability and accuracy, especially in outdoor scenarios. WK-Pnet demonstrates strong robustness, low-latency inference, and high accuracy, making it highly suitable for real-time, resource-constrained mobile and IoT applications.

Recommender systems have been widely used in e-commerce, and re-ranking models are playing an increasingly significant role in the domain, which leverages the inter-item influence and determines the final recommendation lists. Online learning methods keep updating a deployed model with the latest available samples to capture the shifting of the underlying data distribution in e-commerce. However, they depend on the availability of real user feedback, which may be delayed by hours or even days, such as item purchases, leading to a lag in model enhancement. In this paper, we propose a novel extension of online learning methods for re-ranking modeling, which we term LAST, an acronym for Learning At Serving Time. It circumvents the requirement of user feedback by using a surrogate model to provide the instructional signal needed to steer model improvement. Upon receiving an online request, LAST finds and applies a model modification on the fly before generating a recommendation result for the request. The modification is request-specific and transient. It means the modification is tailored to and only to the current request to capture the specific context of the request. After a request, the modification is discarded, which helps to prevent error propagation and stabilizes the online learning procedure since the predictions of the surrogate model may be inaccurate. Most importantly, as a complement to feedback-based online learning methods, LAST can be seamlessly integrated into existing online learning systems to create a more adaptive and responsive recommendation experience. Comprehensive experiments, both offline and online, affirm that LAST outperforms state-of-the-art re-ranking models.

Multi-stage ranking pipelines have become widely used strategies in modern recommender systems, where the final stage aims to return a ranked list of items that balances a number of requirements such as user preference, diversity, novelty etc. Linear scalarization is arguably the most widely used technique to merge multiple requirements into one optimization objective, by summing up the requirements with certain preference weights. Existing final-stage ranking methods often adopt a static model where the preference weights are determined during offline training and kept unchanged during online serving. Whenever a modification of the preference weights is needed, the model has to be re-trained, which is time and resources inefficient. Meanwhile, the most appropriate weights may vary greatly for different groups of targeting users or at different time periods (e.g., during holiday promotions). In this paper, we propose a framework called controllable multi-objective re-ranking (CMR) which incorporates a hypernetwork to generate parameters for a re-ranking model according to different preference weights. In this way, CMR is enabled to adapt the preference weights according to the environment changes in an online manner, without retraining the models. Moreover, we classify practical business-oriented tasks into four main categories and seamlessly incorporate them in a new proposed re-ranking model based on an Actor-Evaluator framework, which serves as a reliable real-world testbed for CMR. Offline experiments based on the dataset collected from Taobao App showed that CMR improved several popular re-ranking models by using them as underlying models. Online A/B tests also demonstrated the effectiveness and trustworthiness of CMR.

Purchase-related micro-behaviors, e.g., favorite, add to cart, read reviews, etc., provide implicit feedback of users' decision-making process. Such informative feedback can lead to fine-grained post-click conversion rate (CVR) modeling of the buying process. However, most existing works on CVR estimation either neglect these informative feedback, or model them as a sequential pattern with Recurrent Neural Networks. We argue such modeling could be inappropriate since different orders of micro-behaviors may represent similar user buying intention, and micro-behaviors often correlate with each other.

Recommender system, as an essential part of modern e-commerce, consists of two fundamental modules, namely Click-Through Rate (CTR) and Conversion Rate (CVR) prediction. While CVR has a direct impact on the purchasing volume, its prediction is well-known challenging due to the Sample Selection Bias (SSB) and Data Sparsity (DS) issues. Although existing methods, typically built on the user sequential behavior path "impression->click->purchase", is effective for dealing with SSB issue, they still struggle to address the DS issue due to rare purchase training samples. Observing that users always take several purchase-related actions after clicking, we propose a novel idea of post-click behavior decomposition. Specifically, disjoint purchase-related Deterministic Action (DAction) and Other Action (OAction) are inserted between click and purchase in parallel, forming a novel user sequential behavior graph "impression->click->D(O)Action->purchase". Defining model on this graph enables to leverage all the impression samples over the entire space and extra abundant supervised signals from D(O)Action, which will effectively address the SSB and DS issues together. To this end, we devise a novel deep recommendation model named Elaborated Entire Space Supervised Multi-task Model (ESM2). According to the conditional probability rule defined on the graph, it employs multi-task learning to predict some decomposed sub-targets in parallel and compose them sequentially to formulate the final CVR. Extensive experiments on both offline and online environments demonstrate the superiority of ESM2 over state-of-the-art models. The source code and dataset will be released.

Post-click conversion rate (CVR) estimation is a critical task in e-commerce recommender systems. This task is deemed quite challenging under industrial setting with two major issues: 1) selection bias caused by user self-selection, and 2) data sparsity due to the rare click events. A successful conversion typically has the following sequential events: ”exposure → click → conversion”. Conventional CVR estimators are trained in the click space, but inference is done in the entire exposure space. They fail to account for the causes of the missing data and treat them as missing at random. Hence, their estimations are highly likely to deviate from the real values by large. In addition, the data sparsity issue can also handicap many industrial CVR estimators which usually have large parameter spaces.

Developing effective and efficient recommendation methods is very challenging for modern e-commerce platforms. Generally speaking, two essential modules named “ClickThrough Rate Prediction” (CTR) and “Conversion Rate Prediction” (CVR) are included, where CVR module is a crucial factor that affects the final purchasing volume directly. However, it is indeed very challenging due to its sparseness nature. In this paper, we tackle this problem by proposing multiLevel Deep Cascade Trees (ldcTree), which is a novel decision tree ensemble approach. It leverages deep cascade structures by stacking Gradient Boosting Decision Trees (GBDT) to effectively learn feature representation. In addition, we propose to utilize the cross-entropy in each tree of the preceding GBDT as the input feature representation for next level GBDT, which has a clear explanation, i.e., a traversal from root to leaf nodes in the next level GBDT corresponds to the combination of certain traversals in the preceding GBDT. The deep cascade structure and the combination rule enable the proposed ldcTree to have a stronger distributed feature representation ability. Moreover, inspired by ensemble learning, we propose an Ensemble ldcTree (E-ldcTree) to encourage the model’s diversity and enhance the representation ability further. Finally, we propose an improved Feature learning method based on EldcTree (F-EldcTree) for taking adequate use of weak and strong correlation features identified by pretrained GBDT models. Experimental results on off-line data set and online deployment demonstrate the effectiveness of the proposed methods.

Capturing users' precise preferences is a fundamental problem in large-scale recommender system. Currently, item-based Collaborative Filtering (CF) methods are common matching approaches in industry. However, they are not effective to model dynamic and evolving preferences of users. In this paper, we propose a new sequential deep matching (SDM) model to capture users' dynamic preferences by combining short-term sessions and long-term behaviors. Compared with existing sequence-aware recommendation methods, we tackle the following two inherent problems in real-world applications: (1) there could exist multiple interest tendencies in one session. (2) long-term preferences may not be effectively fused with current session interests. Long-term behaviors are various and complex, hence those highly related to the short-term session should be kept for fusion. We propose to encode behavior sequences with two corresponding components: multi-head self-attention module to capture multiple types of interests and long-short term gated fusion module to incorporate long-term preferences. Successive items are recommended after matching between sequential user behavior vector and item embedding vectors. Offline experiments on real-world datasets show the superior performance of the proposed SDM. Moreover, SDM has been successfully deployed on online large-scale recommender system at Taobao and achieves improvements in terms of a range of commercial metrics.

Visual analytics is inherently a collaboration between human and computer. However, in current visual analytics systems, the computer has limited means of knowing about its users and their analysis processes. While existing research has shown that a user's interactions with a system reflect a large amount of the user's reasoning process, there has been limited advancement in developing automated, real-time techniques that mine interactions to learn about the user. In this paper, we demonstrate that we can accurately predict a user's task performance and infer some user personality traits by using machine learning techniques to analyze interaction data. Specifically, we conduct an experiment in which participants perform a visual search task, and apply well-known machine learning algorithms to three encodings of the users' interaction data. We achieve, depending on algorithm and encoding, between 62% and 83% accuracy at predicting whether each user will be fast or slow at completing the task. Beyond predicting performance, we demonstrate that using the same techniques, we can infer aspects of the user's personality factors, including locus of control, extraversion, and neuroticism. Further analyses show that strong results can be attained with limited observation time: in one case 95% of the final accuracy is gained after a quarter of the average task completion time. Overall, our findings show that interactions can provide information to the computer about its human collaborator, and establish a foundation for realizing mixed-initiative visual analytics systems.

It is well known that users' behaviors (actions) in a social network are influenced by various factors such as personal interests, social influence, and global trends. However, few publications systematically study how social actions evolve in a dynamic social network and to what extent different factors affect the user actions.

BACKGROUND: Each major protein database uses its own conventions when assigning protein identifiers. Resolving the various, potentially unstable, identifiers that refer to identical proteins is a major challenge. This is a common problem when attempting to unify datasets that have been annotated with proteins from multiple data sources or querying data providers with one flavour of protein identifiers when the source database uses another. Partial solutions for protein identifier mapping exist but they are limited to specific species or techniques and to a very small number of databases. As a result, we have not found a solution that is generic enough and broad enough in mapping scope to suit our needs. RESULTS: We have created the Protein Identifier Cross-Reference (PICR) service, a web application that provides interactive and programmatic (SOAP and REST) access to a mapping algorithm that uses the UniProt Archive (UniParc) as a data warehouse to offer protein cross-references based on 100% sequence identity to proteins from over 70 distinct source databases loaded into UniParc. Mappings can be limited by source database, taxonomic ID and activity status in the source database. Users can copy/paste or upload files containing protein identifiers or sequences in FASTA format to obtain mappings using the interactive interface. Search results can be viewed in simple or detailed HTML tables or downloaded as comma-separated values (CSV) or Microsoft Excel (XLS) files suitable for use in a local database or a spreadsheet. Alternatively, a SOAP interface is available to integrate PICR functionality in other applications, as is a lightweight REST interface. CONCLUSION: We offer a publicly available service that can interactively map protein identifiers and protein sequences to the majority of commonly used protein databases. Programmatic access is available through a standards-compliant SOAP interface or a lightweight REST interface. The PICR interface, documentation and code examples are available at http://www.ebi.ac.uk/Tools/picr.