Recommender systems are fundamental information filtering techniques to recommend content or items that meet users’ personalities and potential needs. As a crucial solution to address the difficulty of user identification and unavailability of historical information, session-based recommender systems provide recommendation services that only rely on users’ behaviors in the current session. However, most existing studies are not well-designed for modeling heterogeneous user behaviors and capturing the relationships between them in practical scenarios. To fill this gap, in this article, we propose a novel graph-based method, namely H eterogeneous I nformation C rossing on G raphs (HICG). HICG utilizes multiple types of user behaviors in the sessions to construct heterogeneous graphs, and captures users’ current interests with their long-term preferences by effectively crossing the heterogeneous information on the graphs. In addition, we also propose an enhanced version, named HICG-CL, which incorporates the contrastive learning (CL) technique to enhance item representation ability. By utilizing the item co-occurrence relationships across different sessions, HICG-CL improves the recommendation performance of HICG. We conduct extensive experiments on three real-world recommendation datasets, and the results verify that (i) HICG achieves state-of-the-art performance by utilizing multiple types of behaviors on the heterogeneous graph. (ii) HICG-CL further significantly improves the recommendation performance of HICG by the proposed contrastive learning module.

Prompt tuning has enhanced the performance of Pre-trained Language Models for multi-task learning in few-shot scenarios. However, existing studies fail to consider that the prompts among different layers in Transformer are different due to the diverse information learned at each layer. In general, the bottom layers in the model tend to capture low-level semantic or structural information, while the upper layers primarily acquire task-specific knowledge. Hence, we propose a novel hierarchical prompt tuning model for few-shot multi-task learning to capture this regularity. The designed model mainly consists of three types of prompts: shared prompts, auto-adaptive prompts, and task-specific prompts. Shared prompts facilitate the sharing of general information across all tasks. Auto-adaptive prompts dynamically select and integrate relevant prompt information from all tasks into the current task. Task-specific prompts concentrate on learning task-specific knowledge. To enhance the model's adaptability to diverse inputs, we introduce deep instance-aware language prompts as the foundation for constructing the above prompts. To evaluate the effectiveness of our proposed method, we conduct extensive experiments on multiple widely-used datasets. The experimental results demonstrate that the proposed method achieves state-of-the-art performance for multi-task learning in few-shot settings and outperforms ChatGPT in the full-data setting.

In online recommendation, financial service, etc., the most common application of multi-task learning (MTL) is the multi-step conversion estimations. A core property of the multi-step conversion is the sequential dependence among tasks. However, most existing works focus far more on the specific post-view click-through rate (CTR) and post-click conversion rate (CVR) estimations, which neglect the generalization of sequential dependence multi-task learning (SDMTL). Additionally, the performance of the SDMTL framework is also deteriorated by the interference derived from implicitly conflict information passing between adjacent tasks. In this paper, a systematic learning paradigm of the SDMTL problem is established for the first time, which can transform the SDMTL problem into a general MTL problem with constraints and be applicable to more general multi-step conversion scenarios with stronger task dependence. Also, the distribution dependence relationship between adjacent task spaces is illustrated from a theoretical point of view. On the other hand, an SDMTL architecture, named Task Aware Feature Extraction (TAFE), is developed to enable dynamic task representation learning from a sample-wise view. TAFE selectively reconstructs the implicit shared information corresponding to each sample case and performs explicit task-specific extraction under dependence constraints. Extensive experiments on offline public and real-world industrial datasets, and online A/B implementations demonstrate the effectiveness and applicability of proposed theoretical and implementation frameworks.

Sequential recommendation (SR) learns users’ preferences by capturing the sequential patterns from users’ behaviors evolution. As discussed in many works, user–item interactions of SR generally present the intrinsic power-law distribution, which can be ascended to hierarchy-like structures. Previous methods usually handle such hierarchical information by making user–item sectionalization empirically under Euclidean space, which may cause distortion of user–item representation in real online scenarios. In this article, we propose a Poincaré-based heterogeneous graph neural network named Poincaré Heterogeneous Graph Neural Networks for Sequential Recommendation (PHGR) to model the sequential pattern information as well as hierarchical information contained in the data of SR scenarios simultaneously. Specifically, for the purpose of explicitly capturing the hierarchical information, we first construct a weighted user–item heterogeneous graph by aliening all the user–item interactions to improve the perception domain of each user from a global view. Then the output of the global representation would be used to complement the local directed item–item homogeneous graph convolution. By defining a novel hyperbolic inner product operator, the global and local graph representation learning are directly conducted in Poincaré ball instead of commonly used projection operation between Poincaré ball and Euclidean space, which could alleviate the cumulative error issue of general bidirectional translation process. Moreover, for the purpose of explicitly capturing the sequential dependency information, we design two types of temporal attention operations under Poincaré ball space. Empirical evaluations on datasets from the public and financial industry show that PHGR outperforms several comparison methods.

In this paper, we present a new cross-architecture contrastive learning (CACL) framework for self-supervised video representation learning. CACL consists of a 3D CNN and a video transformer which are used in parallel to generate diverse positive pairs for contrastive learning. This allows the model to learn strong representations from such diverse yet meaningful pairs. Furthermore, we introduce a temporal self-supervised learning module able to predict an Edit distance explicitly between two video sequences in the temporal order. This enables the model to learn a rich temporal representation that compensates strongly to the video-level representation learned by the CACL. We evaluate our method on the tasks of video retrieval and action recognition on UCF101 and HMDB51 datasets, where our method achieves excellent performance, surpassing the state-of-the-art methods such as VideoMoCo [23] and MoCo+BE [34] by a large margin.

Traditional object detectors employ the dense paradigm of scanning over locations and scales in an image. The recent query-based object detectors break this convention by decoding image features with a set of learnable queries. However, this paradigm still suffers from slow convergence, limited performance, and design complexity of extra networks between backbone and decoder. In this paper, we find that the key to these issues is the adaptability of decoders for casting queries to varying objects. Accordingly, we propose a fast-converging query-based detector, named AdaMixer, by improving the adaptability of query-based decoding processes in two aspects. First, each query adaptively samples features over space and scales based on estimated offsets, which allows AdaMixer to efficiently attend to the coherent regions of objects. Then, we dynamically decode these sampled features with an adaptive MLP-Mixer under the guidance of each query. Thanks to these two critical designs, AdaMixer enjoys architectural simplicity without requiring dense attentional encoders or explicit pyramid networks. On the challenging MS COCO benchmark, AdaMixer with ResNet-50 as the backbone, with 12 training epochs, reaches up to 45.0 AP on the validation set along with 27.9 APs in detecting small objects. With the longer training scheme, AdaMixer with ResNeXt-101-DCN and Swin-S reaches 49.5 and 51.3 AP. Our work sheds light on a simple, accurate, and fast converging architecture for query-based object detectors. The code is made available at https://github.com/MCG-NJU/AdaMixer.