📖 Overview

This project explores the unification of multiple recommendation tasks within a single, end-to-end generative model.

• IntTravel: Our foundational work that introduced a large-scale, real-world dataset and a generative framework for integrated multi-task travel recommendation.
• IntHQ: The advanced successor to IntTravel, which identifies and resolves the "threefold collapse" in generative multi-task models with a novel architecture.

IntTravel: A Large-Scale Real-World Dataset

We introduce IntTravel, the first large-scale public dataset for integrated travel recommendation, including 4.1 billion interactions from 163 million users with 7.3 million POIs. Built upon this dataset, we introduce an end-to-end, decoder-only generative framework for multi-task recommendation.

All data are collected from a leading provider of digital map, navigation and real-time traffic information in China. Here is a simple dataset in data_process/raw_data and a more comprehensive dataset in Hugging Face. The code in data_process demonstrates how to construct the input sequence of the model and the labels for all tasks based on the original data.

Information of POIs

The IntTravel dataset contains 7,291,872 POIs (Point of Interests) distributed across several major cities in China. Each POI is described by the following fields:

Field	Description
POI ID	A unique identifier for each Point of Interest.
Normalized score	A 0-1 score reflecting the overall popularity of the POI.
Geographic ID	Identifier for the POI's geographic block. Same GIDs indicate geographical proximity.
Category ID	A numerical identifier for the Point of Interest's category.
Administrative Region ID	The identifier for the administrative region of the POI.
Coordinates	The spatial coordinates of the POI on a 2D plane.

User Profiles

The IntTravel dataset contains 162,815,861 users, each described by the following fields:

Field	Description
User ID	A unique identifier assigned to each user.
Profile Feature 1	The first profile feature.
...	...
Profile Feature 6	The sixth profile feature.

User Interactions

The IntTravel dataset includes 4,129,827,011 user interaction events. Each event is characterized by the following fields:

Field	Description
User ID	A unique identifier for the user who performed the interaction.
Timestamp	The time of the user interaction, recorded in milliseconds.
Action Type	A numerical ID representing the type of user behavior (e.g., click).
POI ID	The identifier of the Point of Interest involved in the interaction.
Geographic ID	The geographic block ID where the user was during the interaction.
Administrative Region ID	The administrative region ID where the user was during the interaction.
Weather	A numerical ID representing the weather condition during the interaction.
Travel Mode	A numerical ID for the user's chosen travel mode.
Via POI ID	The identifier for a way-point POI added by the user.

IntTravel: Generative Framework for Integrated Multi-Task Travel Recommendation

IntTravel incorporates information preservation, selection, and factorization to balance task collaboration with specialized differentiation, yielding substantial performance gains. IntTravel has been successfully deployed on Amap serving hundreds of millions of users.

IntTravel is the first multi-task solution for generative recommendation. We propose a bottom-up multi-task method to handle multiple tasks within a single generative model. The approach comprises three modules:

• Task-Guided Information Persistence (TIP) ensures maximum propagation of task-relevant information in the decoder.
• Task-Specific Selective Gating (TSG) enables each task to filter useful information from the decoder's output.
• Task-Aware Scenario Factorization (TSF) empowers each task to factorize its output based on specific scenarios.

IntHQ: Task-Interactive Hierarchical Query on Dual-Stream Representations for Generative Recommendation

Multi-task learning over heterogeneous data is fundamental to modern recommendation, while generative models are emerging as the backbone of next-generation recommenders. However, the integration of multi-task learning into the generative paradigm remains largely unexplored. Existing multi-task recommenders, in both discriminative and generative paradigms, extract task-relevant features from a single task-agnostic representation and wire tasks into a predefined conversion funnel. We show that this scheme is inherently prone to a threefold collapse.

• Source collapse, where task-specific signals are injected late and diluted in the shared latent space.
• Relational collapse, where task dependencies are either implicitly absorbed by the backbone or statically fixed by predefined funnels.
• Hierarchical collapse, where tasks depend on features at different scales and shift across training stages.

IntHQ is a multi-task generative recommender with three components, each alleviating one collapse:

• Dual-Stream Decoupling (DSD) injects task identity into the computation stream early and separates the shared context stream from the task-specific stream, alleviating signal dilution.

• Task-Interactive Modeling (TIM) replaces the predefined funnel with explicit cross-task interaction, letting each task condition on the realized outcomes of its predecessors with learned, input-adaptive strength.

• Hierarchical Querying (HQ) lets each task gather multi-scale information across different layers at different training stages.

In offline evaluations, IntHQ consistently outperforms competitive encoder backbones under four representative task-head configurations. Deployed in production on Amap, serving hundreds of millions of users for travel recommendation, IntHQ yields a 1.60% relative UVCTR lift.

📚 Citation

If you find our papers and code helpful for your research, please consider starring our repository ⭐ and citing our work ✏️.

@article{yan2026inttravel,
  title={IntTravel: A Real-World Dataset and Generative Framework for Integrated Multi-Task Travel Recommendation},
  author={Yan, Huimin and Xu, Longfei and Sun, Junjie and Liu, Zheng and Luo, Wei and Liu, Kaikui and Chu, Xiangxiang},
  journal={arXiv preprint arXiv:2602.11664},
  year={2026}
}
@article{sun2026inthq,
  title={IntHQ: Task-Interactive Hierarchical Query on Dual-Stream Representations for Generative Recommendation},
  author={Sun, Junjie and Xu, Longfei and Yan, Huimin and Luo, Wei and Liu, Kaikui and Chu, Xiangxiang},
  journal={arXiv preprint arXiv:2608.09634},
  year={2026}
}