Saved in:
| Main Authors: | , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2503.06706 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913891437510656 |
|---|---|
| author | Zhang, Ming Wang, Yuhui Shen, Yujiong Yang, Tingyi Jiang, Changhao Wu, Yilong Dou, Shihan Chen, Qinhao Xi, Zhiheng Zhang, Zhihao Dong, Yi Wang, Zhen Fei, Zhihui Wan, Mingyang Liang, Tao Ma, Guojun Zhang, Qi Gui, Tao Huang, Xuanjing |
| author_facet | Zhang, Ming Wang, Yuhui Shen, Yujiong Yang, Tingyi Jiang, Changhao Wu, Yilong Dou, Shihan Chen, Qinhao Xi, Zhiheng Zhang, Zhihao Dong, Yi Wang, Zhen Fei, Zhihui Wan, Mingyang Liang, Tao Ma, Guojun Zhang, Qi Gui, Tao Huang, Xuanjing |
| contents | Process-driven dialogue systems, which operate under strict predefined process constraints, are essential in customer service and equipment maintenance scenarios. Although Large Language Models (LLMs) have shown remarkable progress in dialogue and reasoning, they still struggle to solve these strictly constrained dialogue tasks. To address this challenge, we construct Process Flow Dialogue (PFDial) dataset, which contains 12,705 high-quality Chinese dialogue instructions derived from 440 flowcharts containing 5,055 process nodes. Based on PlantUML specification, each UML flowchart is converted into atomic dialogue units i.e., structured five-tuples. Experimental results demonstrate that a 7B model trained with merely 800 samples, and a 0.5B model trained on total data both can surpass 90% accuracy. Additionally, the 8B model can surpass GPT-4o up to 43.88% with an average of 11.00%. We further evaluate models' performance on challenging backward transitions in process flows and conduct an in-depth analysis of various dataset formats to reveal their impact on model performance in handling decision and sequential branches. The data is released in https://github.com/KongLongGeFDU/PFDial. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_06706 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | PFDial: A Structured Dialogue Instruction Fine-tuning Method Based on UML Flowcharts Zhang, Ming Wang, Yuhui Shen, Yujiong Yang, Tingyi Jiang, Changhao Wu, Yilong Dou, Shihan Chen, Qinhao Xi, Zhiheng Zhang, Zhihao Dong, Yi Wang, Zhen Fei, Zhihui Wan, Mingyang Liang, Tao Ma, Guojun Zhang, Qi Gui, Tao Huang, Xuanjing Computation and Language Artificial Intelligence Machine Learning Process-driven dialogue systems, which operate under strict predefined process constraints, are essential in customer service and equipment maintenance scenarios. Although Large Language Models (LLMs) have shown remarkable progress in dialogue and reasoning, they still struggle to solve these strictly constrained dialogue tasks. To address this challenge, we construct Process Flow Dialogue (PFDial) dataset, which contains 12,705 high-quality Chinese dialogue instructions derived from 440 flowcharts containing 5,055 process nodes. Based on PlantUML specification, each UML flowchart is converted into atomic dialogue units i.e., structured five-tuples. Experimental results demonstrate that a 7B model trained with merely 800 samples, and a 0.5B model trained on total data both can surpass 90% accuracy. Additionally, the 8B model can surpass GPT-4o up to 43.88% with an average of 11.00%. We further evaluate models' performance on challenging backward transitions in process flows and conduct an in-depth analysis of various dataset formats to reveal their impact on model performance in handling decision and sequential branches. The data is released in https://github.com/KongLongGeFDU/PFDial. |
| title | PFDial: A Structured Dialogue Instruction Fine-tuning Method Based on UML Flowcharts |
| topic | Computation and Language Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2503.06706 |