Learning transferable representations from unlabeled time series is crucial for improving performance in data-scarce classification. Existing self-supervised methods often operate at the point level and rely on unidirectional encoding, leading to low semantic density and a mismatch between pre-training and downstream optimization. In this paper, we propose TimeMAE, a self-supervised framework that reformulates masked modeling for time series via semantic unit elevation and decoupled representation learning. Instead of modeling individual time steps, TimeMAE segments time series into non-overlapping sub-series to form semantically enriched units, enabling more informative masked reconstruction while reducing computational cost. To address the representation discrepancy introduced by masking, we design a decoupled masked autoencoder that separately encodes visible and masked regions, avoiding artificial masked tokens in the main encoder. To guide pre-training, we introduce two complementary objectives: masked codeword classification, which discretizes sub-series semantics via a learned tokenizer and masked representation regression, which aligns continuous representations through a momentum-updated target encoder. Extensive experiments on five datasets demonstrate that TimeMAE outperforms competitive baselines, particularly in label-scarce scenarios and transfer learning scenarios. Our codes are publicly available at https://github.com/Mingyue-Cheng/TimeMAE.

Search-based recommendation is one of the most critical application scenarios in e-commerce platforms. Users' complex search contexts—such as spatiotemporal factors, historical interactions, and current query's information—constitute an essential part of their decision-making, reflecting implicit preferences that complement explicit query terms. Modeling such rich contextual signals and their intricate associations with candidate items remains a key challenge. Although numerous efforts have been devoted to building more effective search methods, existing approaches still show limitations in integrating contextual information, which hinders their ability to fully capture user intent. To address these challenges, we propose a context-aware reasoning-enhanced generative search framework for better understanding the complicated context. Specifically, the framework first unifies heterogeneous user and item contexts into textual representations or text-based semantic identifiers and aligns them. To overcome the lack of explicit reasoning trajectories, we introduce a self-evolving post-training paradigm that iteratively combines supervised fine-tuning and reinforcement learning to progressively enhance the model's reasoning capability. In addition, we identify potential biases in existing RL algorithms when applied to search scenarios and present a debiased variant of GRPO to improve ranking performance. Extensive experiments on search log data collected from a real-world e-commerce platform demonstrate that our approach achieves superior performance compared with strong baselines, validating its effectiveness for search-based recommendation.

For the advancement of time series classification, we can summarize that most existing methods adopt a common learning-to-classify paradigm - a classifier model tries to learn the relation between sequence inputs and target label encoded by one-hot distribution. Although effective, this paradigm conceals two inherent limitations: (1) one-hot distribution fails to reflect the comparability and similarity between labels, and (2) it is difficult to learn transferable representation across domains. In this work, we propose InstructTime, a novel attempt to reshape time series classification as a learning-to-generate paradigm. Relying on the generative capacity of the pre-trained language model, the core idea is to formulate the classification of time series as a multimodal understanding task. Specifically, firstly, a time series discretization module is designed to convert continuous inputs into a sequence of discrete tokens to solve the inconsistency issue across modality data. Secondly, we introduce an alignment projected layer before feeding the transformed token of time series into language models. Thirdly, prior to fine-tuning the language model for the target domain, it is essential to emphasize the necessity of auto-regressive pre-training across various modality inputs. Finally, extensive experimentation are conducted on several prevalent public benchmark datasets, indicating the superior performance of the InstructTime. Our code is at https://github.com/Mingyue-Cheng/InstructTime.

Sequential recommender systems offer personalized suggestions by modeling users' interactions chronologically to capture dynamic user interest. Existing approaches typically fail to adequately describe the dynamics of the entire recommender system, including shifts in both user interest and item availability. To address this, we propose a simple yet effective framework with three key perspectives, tailored to the dynamics of recommender system by fully exploiting the time information. Firstly, we propose a dynamic candidate set construction approach to prevent the model from learning future interactions. Secondly, assuming that user behaviors remain consistent over short terms but may evolve over long terms, we employ a interval-weighted optimization target to model the correlation of users' historical interactions. Finally, we introduce a specialized time-aware attention module to enhance recommendations within specific temporal contexts. Extensive experiments demonstrate the effectiveness and generalizability of our framework. We make our codes publicly available.

Recent efforts have been dedicated to enhancing time series forecasting accuracy by introducing advanced network architectures and self-supervised pretraining strategies. Nevertheless, existing approaches still exhibit two critical drawbacks. Firstly, these methods often rely on a single dataset for training, limiting the model's generalizability due to the restricted scale of the training data. Secondly, the one-step generation schema is widely followed, which necessitates a customized forecasting head and overlooks the temporal dependencies in the output series, and also leads to increased training costs under different horizon length settings.

The fundamental task of intelligent educational systems is to offer adaptive learning services to students, such as exercise recommendations and computerized adaptive testing. However, optimizing required models in these systems would always encounter the collection difficulty of high-quality interaction data in practice. Therefore, establishing a student simulator is of great value since it can generate valid interactions to help optimize models. Existing advances have achieved success but generally suffer from exposure bias and overlook long-term intentions. To tackle these problems, we propose a novel Direct-Adversarial Imitation Student Simulator (DAISim) by formulating it as a Markov Decision Process (MDP), which unifies the workflow of the simulator in training and generating to alleviate the exposure bias and single-step optimization problems. To construct the intentions underlying the complex student interactions, we first propose a direct imitation strategy to mimic the interactions with a simple reward function. Then, we propose an adversarial imitation strategy to learn a rational distribution with the reward given by a parameterized discriminator. Furthermore, we optimize the discriminator in adversarial imitation in a pairwise manner, and the theoretical analysis shows that the pairwise discriminator would improve the generation quality. We conduct extensive experiments on real-world datasets, where the results demonstrate that our DAISim can simulate high-quality student interactions whose distribution is close to real distribution and can promote several downstream services.

Deep learning-based algorithms, e.g., convolutional networks, have significantly facilitated multivariate time series classification (MTSC) task. Nevertheless, they suffer from the limitation in modeling long-range dependence due to the nature of convolution operations. Recent advancements have shown the potential of transformers to capture long-range dependence. However, it would incur severe issues, such as fixed scale representations, temporal-invariant and quadratic time complexity, with transformers directly applicable to the MTSC task because of the distinct properties of time series data. To tackle these issues, we propose FormerTime, an hierarchical representation model for improving the classification capacity for the MTSC task. In the proposed FormerTime, we employ a hierarchical network architecture to perform multi-scale feature maps. Besides, a novel transformer encoder is further designed, in which an efficient temporal reduction attention layer and a well-informed contextual positional encoding generating strategy are developed. To sum up, FormerTime exhibits three aspects of merits: (1) learning hierarchical multi-scale representations from time series data, (2) inheriting the strength of both transformers and convolutional networks, and (3) tacking the efficiency challenges incurred by the self-attention mechanism. Extensive experiments performed on 10 publicly available datasets from UEA archive verify the superiorities of the FormerTime compared to previous competitive baselines.

Deep learning has brought significant breakthroughs in sequential recommendation (SR) for capturing dynamic user interests. A series of recent research revealed that models with more parameters usually achieve optimal performance for SR tasks, inevitably resulting in great challenges for deploying them in real systems. Following the simple assumption that light networks might already suffice for certain users, in this work, we propose CANet, a conceptually simple yet very scalable framework for assigning adaptive network architecture in an input-dependent manner to reduce unnecessary computation. The core idea of CANet is to route the input user behaviors with a light-weighted router module. Specifically, we first construct the routing space with various submodels parameterized in terms of multiple model dimensions such as the number of layers, hidden size and embedding size. To avoid extra storage overhead of the routing space, we employ a weight-slicing schema to maintain all the submodels in exactly one network. Furthermore, we leverage several solutions to solve the discrete optimization issues caused by the router module. Thanks to them, CANet could adaptively adjust its network architecture for each input in an end-to-end manner, in which the user preference can be effectively captured. To evaluate our work, we conduct extensive experiments on benchmark datasets. Experimental results show that CANet reduces computation by 55 ~ 65% while preserving the accuracy of the original model. Our codes are available at https://github.conicantnamemyself/CANet.

Recent years have witnessed great success in deep learning-based sequential recommendation (SR), which can provide more timely and accurate recommendations. One of the most effective deep SR architectures is to stack high-performance residual blocks, e.g., prevalent self-attentive and convolutional operations, for capturing long- and short-range dependence of sequential behaviors. By carefully revisiting previous models, we observe: 1) simple architecture modification of gating each residual connection can help us train deeper SR models and yield significant improvements; 2) compared with self-attention mechanism, stacking of convolution layers also can cover each item of the whole sequential behaviors and achieve competitive or even superior performance.