osunlp/Multimodal-Mind2Web

Dataset Summary

Multimodal-Mind2Web is the multimodal version of Mind2Web, a dataset for developing and evaluating generalist agents for the web that can follow language instructions to complete complex tasks on any website. In this dataset, we align each HTML document in the dataset with its corresponding webpage screenshot image from the Mind2Web raw dump. This multimodal version addresses the inconvenience of loading images from the ~300GB Mind2Web Raw Dump.

Dataset Structure

Data Splits

• train: 7775 actions from 1009 tasks.
• test_task: 1339 actions from 177 tasks. Tasks from the same website are seen during training.
• test_website: 1019 actions from 142 tasks. Websites are not seen during training.
• test_domain: 4060 actions from 694 tasks. Entire domains are not seen during training.

The train set may include some screenshot images not properly rendered caused by rendering issues during Mind2Web annotation. The three test splits (test_task, test_website, test_domain) have undergone human verification to confirm element visibility and correct rendering for action prediction.

Data Fields

Each line in the dataset is an action consisting of screenshot image, HTML text and other fields required for action prediction, for the convenience of inference.

• "annotation_id" (str): unique id for each task
• "website" (str): website name
• "domain" (str): website domain
• "subdomain" (str): website subdomain
• "confirmed_task" (str): task description
• "action_reprs" (list[str]): human readable string representation of the action sequence
• "screenshot" (str): path to the webpage screenshot image corresponding to the HTML.
• "action_uid" (str): unique id for each action (step)
• "raw_html" (str): raw html of the page before the action is performed
• "cleaned_html" (str): cleaned html of the page before the action is performed
• "operation" (dict): operation to perform
  • "op" (str): operation type, one of CLICK, TYPE, SELECT
  • "original_op" (str): original operation type, contain additional HOVER and ENTER that are mapped to CLICK, not used
  • "value" (str): optional value for the operation, e.g., text to type, option to select
• "pos_candidates" (list[dict]): ground truth elements. Here we only include positive elements that exist in "cleaned_html" after our preprocessing, so "pos_candidates" might be empty. The original labeled element can always be found in the "raw_html".
  • "tag" (str): tag of the element
  • "is_original_target" (bool): whether the element is the original target labeled by the annotator
  • "is_top_level_target" (bool): whether the element is a top level target find by our algorithm. please see the paper for more details.
  • "backend_node_id" (str): unique id for the element
  • "attributes" (str): serialized attributes of the element, use json.loads to convert back to dict
• "neg_candidates" (list[dict]): other candidate elements in the page after preprocessing, has similar structure as "pos_candidates"

Disclaimer

This dataset was collected and released solely for research purposes, with the goal of making the web more accessible via language technologies. The authors are strongly against any potential harmful use of the data or technology to any party.

Citation Information

@article{zheng2024seeact,
  title={GPT-4V(ision) is a Generalist Web Agent, if Grounded},
  author={Boyuan Zheng and Boyu Gou and Jihyung Kil and Huan Sun and Yu Su},
  journal={arXiv preprint arXiv:2401.01614},
  year={2024},
}

@inproceedings{deng2023mindweb,
  title={Mind2Web: Towards a Generalist Agent for the Web},
  author={Xiang Deng and Yu Gu and Boyuan Zheng and Shijie Chen and Samuel Stevens and Boshi Wang and Huan Sun and Yu Su},
  booktitle={Thirty-seventh Conference on Neural Information Processing Systems},
  year={2023},
  url={https://openreview.net/forum?id=kiYqbO3wqw}
}