GPlan: Generative Spatiotemporal Intent Sequence Recommendation via Implicit Reasoning in Amap

AMAP, Alibaba Group

📋 Overview

Progressive Implicit CoT distillation training framework of GPlan. Uses curriculum learning to compress structured CoT texts into special think tokens epoch by epoch, distilling implicit reasoning capabilities.

📊GSISR Dataset

The GSISR dataset is collected from Amap and provided in data_process/dataset/. All data have been anonymized to protect privacy — original feature names, POI identifiers, and user identifiers have been replaced with generic placeholders.

File	Description	Num
data_process/dataset/train.csv	Training set	100,000
data_process/dataset/test.csv	Test set	1,000

User Profiles and Behavior History

Each user is described by 14 anonymized profile features. All categorical values have been mapped to numerical IDs.

Field	Description
User ID	A unique numerical identifier for each user.
Profile Feature 1–14	Anonymized user profile attributes.
Short-term Behavior Seq	Anonymized short-term behavior sequence. POI names and behavior types (e.g., click) are replaced with numerical IDs (p_, act_).
Long-term Behavior	Anonymized long-term behavior feature. Original values are replaced with numerical IDs.

Context Information

Field	Description
Current Time	Time of the request.
Weekend Flag	Whether the current day is a weekend (0/1).
Holiday Flag	Whether the current day is a holiday (0/1).
Current City & District	The city and district where the user is located.
Current POI Name	The name of the user's current Point of Interest (mapped to ID, p_).
Current POI Category	The tag of the current POI.

Trigger Events

Each request includes 7 trigger event features that capture the user's immediate intent signals.

Field	Description
Trigger 1–7	Anonymized event trigger features. Original event types and descriptions are replaced with numerical IDs or kept as timestamps.

Labels

Each label is an intent sequence — a JSON array of tool-calling intents representing the recommendation. Each intent includes a tool name and associated parameters selected from a predefined intent library:

[
  {"工具名称": "tool_5", "起始位置": "当前位置", "空间范围": "附近", "tag": "美食"},
  {"工具名称": "tool_2", "起始位置": "当前位置", "终点位置": "家"},
  ...
]

The intent library includes 10 tool types covering scenarios such as ride-hailing, navigation, transit, POI recommendation, order reminders, weather queries, etc.

Note: The CoT (Chain-of-Thought) reasoning text used during training is not included in the public dataset release. Only the final intent sequences are provided as labels.

Preparing CoT Text for Training

The training pipeline relies on the raw_labels column in the CSV, which must contain both the CoT reasoning text and the intent sequence JSON. Since the public dataset only provides the final intent sequences, you need to generate the CoT text yourself before training.

Step 1: Generate CoT text using a large language model

Use a capable LLM (e.g., GPT, Qwen) to generate structured CoT reasoning for each sample. The CoT must follow this XML-tagged format:

<THOUGHT>
<CONTEXT>Briefly analyze the current context and the user's potential needs</CONTEXT>
<STRATEGY>Based on the context analysis, devise the core strategy for the plan</STRATEGY>
<STEP_1>Focus on analyzing the primary and most crucial intent</STEP_1>
...
<STEP_n>Explain why the n-th intent is recommended</STEP_n>
</THOUGHT>

The number of <STEP_n> tags must equal the number of intents in the JSON array.

Step 2: Combine CoT and intent JSON into raw_labels

Concatenate the CoT text and the intent JSON array into a single string, and write it to the raw_labels column of your CSV:

<THOUGHT><CONTEXT>...</CONTEXT><STRATEGY>...</STRATEGY><STEP_1>...</STEP_1><STEP_2>...</STEP_2><STEP_3>...</STEP_3></THOUGHT>[{"工具名称":"tool_5","起始位置":"当前位置","空间范围":"附近","tag":"美食"},{"工具名称":"tool_2","起始位置":"当前位置","终点位置":"家"},{"工具名称":"tool_7","tag":"景点"}]

The training collator (ProgressiveCotDistillCollater) will automatically parse this field, extract the CoT and JSON parts, and apply progressive distillation during training.

🚀 Quick Start

Training

pip install -r requirements.txt
bash finetune.sh

Testing

bash test.sh

📁 Project Structure

├── finetune.py                         # Training script (WeightedLossTrainer + SyncEpochCallback)
├── finetune.sh                         # Training launch script
├── test.py                             # Test script (Accuracy, Edit Similarity, NDCG@3)
├── test.sh                             # Test launch script
├── data_process/
│   ├── dataset/
│   │   ├── train.csv                   # Training dataset (anonymized)
│   │   └── test.csv                    # Test dataset (anonymized)
│   ├── collate_fns.py                  # Data collator (progressive distillation)
│   └── data_loader.py                  # CSV data loading
├── utils.py                            # Utility functions and argument definitions
├── add_tokens/extended_cot_vocabs.json  # CoT special token vocabulary
├── config/ds_z3_bf16.json              # DeepSpeed ZeRO-3 configuration
└── requirements.txt