Datasets:
Tasks:
Text Ranking
Modalities:
Text
Formats:
parquet
Languages:
English
Size:
10M - 100M
ArXiv:
License:
metadata
annotations_creators:
- expert-annotated
language:
- eng
license: cc-by-nc-sa-4.0
multilinguality: monolingual
task_categories:
- text-ranking
task_ids:
- conversational
- utterance-retrieval
dataset_info:
- config_name: corpus
features:
- name: _id
dtype: string
- name: text
dtype: string
- name: title
dtype: string
splits:
- name: validation
num_bytes: 115334059
num_examples: 316508
- name: test
num_bytes: 145478865
num_examples: 405972
- name: test_iid
num_bytes: 147102753
num_examples: 405972
- name: test_geo
num_bytes: 414044181
num_examples: 1150781
- name: test_vis
num_bytes: 597769653
num_examples: 1606858
- name: test_web
num_bytes: 309445145
num_examples: 834175
- name: test_cat
num_bytes: 451206151
num_examples: 1258191
download_size: 368435122
dataset_size: 2180380807
- config_name: default
features:
- name: query-id
dtype: string
- name: corpus-id
dtype: string
- name: score
dtype: int64
splits:
- name: validation
num_bytes: 21945765
num_examples: 316508
- name: test
num_bytes: 23275328
num_examples: 405972
- name: test_iid
num_bytes: 26523104
num_examples: 405972
- name: test_geo
num_bytes: 76517975
num_examples: 1150781
- name: test_vis
num_bytes: 106957051
num_examples: 1606858
- name: test_web
num_bytes: 55262380
num_examples: 834175
- name: test_cat
num_bytes: 83397445
num_examples: 1258191
download_size: 39110732
dataset_size: 393879048
- config_name: queries
features:
- name: _id
dtype: string
- name: text
dtype: string
splits:
- name: validation
num_bytes: 2183028
num_examples: 1301
- name: test
num_bytes: 2510614
num_examples: 1438
- name: test_iid
num_bytes: 2516366
num_examples: 1438
- name: test_geo
num_bytes: 8709463
num_examples: 4916
- name: test_vis
num_bytes: 9363511
num_examples: 5298
- name: test_web
num_bytes: 5846915
num_examples: 3144
- name: test_cat
num_bytes: 7783266
num_examples: 3560
download_size: 4113384
dataset_size: 38913163
- config_name: top_ranked
features:
- name: query-id
dtype: string
- name: corpus-ids
sequence: string
splits:
- name: validation
num_bytes: 12116636
num_examples: 1301
- name: test
num_bytes: 13089849
num_examples: 1438
- name: test_iid
num_bytes: 14719489
num_examples: 1438
- name: test_geo
num_bytes: 42415891
num_examples: 4916
- name: test_vis
num_bytes: 59241856
num_examples: 5298
- name: test_web
num_bytes: 30633101
num_examples: 3144
- name: test_cat
num_bytes: 46195624
num_examples: 3560
download_size: 27777115
dataset_size: 218412446
configs:
- config_name: corpus
data_files:
- split: validation
path: corpus/validation-*
- split: test
path: corpus/test-*
- split: test_iid
path: corpus/test_iid-*
- split: test_geo
path: corpus/test_geo-*
- split: test_vis
path: corpus/test_vis-*
- split: test_web
path: corpus/test_web-*
- split: test_cat
path: corpus/test_cat-*
- config_name: default
data_files:
- split: validation
path: data/validation-*
- split: test
path: data/test-*
- split: test_iid
path: data/test_iid-*
- split: test_geo
path: data/test_geo-*
- split: test_vis
path: data/test_vis-*
- split: test_web
path: data/test_web-*
- split: test_cat
path: data/test_cat-*
- config_name: queries
data_files:
- split: validation
path: queries/validation-*
- split: test
path: queries/test-*
- split: test_iid
path: queries/test_iid-*
- split: test_geo
path: queries/test_geo-*
- split: test_vis
path: queries/test_vis-*
- split: test_web
path: queries/test_web-*
- split: test_cat
path: queries/test_cat-*
- config_name: top_ranked
data_files:
- split: validation
path: top_ranked/validation-*
- split: test
path: top_ranked/test-*
- split: test_iid
path: top_ranked/test_iid-*
- split: test_geo
path: top_ranked/test_geo-*
- split: test_vis
path: top_ranked/test_vis-*
- split: test_web
path: top_ranked/test_web-*
- split: test_cat
path: top_ranked/test_cat-*
tags:
- mteb
- text
WebLINX is a large-scale benchmark of 100K interactions across 2300 expert demonstrations of conversational web navigation. The reranking task focuses on finding relevant elements at every given step in the trajectory.
| Task category | t2t |
| Domains | Academic, Web, Written |
| Reference | https://mcgill-nlp.github.io/weblinx |
How to evaluate on this task
You can evaluate an embedding model on this dataset using the following code:
import mteb
task = mteb.get_tasks(["WebLINXCandidatesReranking"])
evaluator = mteb.MTEB(task)
model = mteb.get_model(YOUR_MODEL)
evaluator.run(model)
To learn more about how to run models on mteb task check out the GitHub repitory.
Citation
If you use this dataset, please cite the dataset as well as mteb, as this dataset likely includes additional processing as a part of the MMTEB Contribution.
@misc{lù2024weblinx,
archiveprefix = {arXiv},
author = {Xing Han Lù and Zdeněk Kasner and Siva Reddy},
eprint = {2402.05930},
primaryclass = {cs.CL},
title = {WebLINX: Real-World Website Navigation with Multi-Turn Dialogue},
year = {2024},
}
@article{enevoldsen2025mmtebmassivemultilingualtext,
title={MMTEB: Massive Multilingual Text Embedding Benchmark},
author={Kenneth Enevoldsen and Isaac Chung and Imene Kerboua and Márton Kardos and Ashwin Mathur and David Stap and Jay Gala and Wissam Siblini and Dominik Krzemiński and Genta Indra Winata and Saba Sturua and Saiteja Utpala and Mathieu Ciancone and Marion Schaeffer and Gabriel Sequeira and Diganta Misra and Shreeya Dhakal and Jonathan Rystrøm and Roman Solomatin and Ömer Çağatan and Akash Kundu and Martin Bernstorff and Shitao Xiao and Akshita Sukhlecha and Bhavish Pahwa and Rafał Poświata and Kranthi Kiran GV and Shawon Ashraf and Daniel Auras and Björn Plüster and Jan Philipp Harries and Loïc Magne and Isabelle Mohr and Mariya Hendriksen and Dawei Zhu and Hippolyte Gisserot-Boukhlef and Tom Aarsen and Jan Kostkan and Konrad Wojtasik and Taemin Lee and Marek Šuppa and Crystina Zhang and Roberta Rocca and Mohammed Hamdy and Andrianos Michail and John Yang and Manuel Faysse and Aleksei Vatolin and Nandan Thakur and Manan Dey and Dipam Vasani and Pranjal Chitale and Simone Tedeschi and Nguyen Tai and Artem Snegirev and Michael Günther and Mengzhou Xia and Weijia Shi and Xing Han Lù and Jordan Clive and Gayatri Krishnakumar and Anna Maksimova and Silvan Wehrli and Maria Tikhonova and Henil Panchal and Aleksandr Abramov and Malte Ostendorff and Zheng Liu and Simon Clematide and Lester James Miranda and Alena Fenogenova and Guangyu Song and Ruqiya Bin Safi and Wen-Ding Li and Alessia Borghini and Federico Cassano and Hongjin Su and Jimmy Lin and Howard Yen and Lasse Hansen and Sara Hooker and Chenghao Xiao and Vaibhav Adlakha and Orion Weller and Siva Reddy and Niklas Muennighoff},
publisher = {arXiv},
journal={arXiv preprint arXiv:2502.13595},
year={2025},
url={https://arxiv.org/abs/2502.13595},
doi = {10.48550/arXiv.2502.13595},
}
@article{muennighoff2022mteb,
author = {Muennighoff, Niklas and Tazi, Nouamane and Magne, Lo{\"\i}c and Reimers, Nils},
title = {MTEB: Massive Text Embedding Benchmark},
publisher = {arXiv},
journal={arXiv preprint arXiv:2210.07316},
year = {2022}
url = {https://arxiv.org/abs/2210.07316},
doi = {10.48550/ARXIV.2210.07316},
}
Dataset Statistics
Dataset Statistics
The following code contains the descriptive statistics from the task. These can also be obtained using:
import mteb
task = mteb.get_task("WebLINXCandidatesReranking")
desc_stats = task.metadata.descriptive_stats
{
"validation": {
"num_samples": 317809,
"number_of_characters": 102848781,
"num_documents": 316508,
"min_document_length": 152,
"average_document_length": 318.17634941296905,
"max_document_length": 1605,
"unique_documents": 316508,
"num_queries": 1301,
"min_query_length": 142,
"average_query_length": 1647.5180630284397,
"max_query_length": 9356,
"unique_queries": 1301,
"none_queries": 0,
"num_relevant_docs": 316508,
"min_relevant_docs_per_query": 21,
"average_relevant_docs_per_query": 1.01076095311299,
"max_relevant_docs_per_query": 945,
"unique_relevant_docs": 316508,
"num_instructions": null,
"min_instruction_length": null,
"average_instruction_length": null,
"max_instruction_length": null,
"unique_instructions": null,
"num_top_ranked": 1301,
"min_top_ranked_per_query": 21,
"average_top_ranked_per_query": 243.2805534204458,
"max_top_ranked_per_query": 945
},
"test_iid": {
"num_samples": 407410,
"number_of_characters": 131631330,
"num_documents": 405972,
"min_document_length": 153,
"average_document_length": 318.135696550501,
"max_document_length": 1471,
"unique_documents": 405972,
"num_queries": 1438,
"min_query_length": 173,
"average_query_length": 1722.6321279554938,
"max_query_length": 10467,
"unique_queries": 1438,
"none_queries": 0,
"num_relevant_docs": 405972,
"min_relevant_docs_per_query": 15,
"average_relevant_docs_per_query": 1.0528511821974966,
"max_relevant_docs_per_query": 1149,
"unique_relevant_docs": 405972,
"num_instructions": null,
"min_instruction_length": null,
"average_instruction_length": null,
"max_instruction_length": null,
"unique_instructions": null,
"num_top_ranked": 1438,
"min_top_ranked_per_query": 15,
"average_top_ranked_per_query": 282.317107093185,
"max_top_ranked_per_query": 1149
},
"test_cat": {
"num_samples": 1261751,
"number_of_characters": 402615943,
"num_documents": 1258191,
"min_document_length": 156,
"average_document_length": 313.91351392594606,
"max_document_length": 1590,
"unique_documents": 1258191,
"num_queries": 3560,
"min_query_length": 161,
"average_query_length": 2149.6587078651687,
"max_query_length": 8502,
"unique_queries": 3560,
"none_queries": 0,
"num_relevant_docs": 1258191,
"min_relevant_docs_per_query": 14,
"average_relevant_docs_per_query": 1.0016853932584269,
"max_relevant_docs_per_query": 1245,
"unique_relevant_docs": 1258191,
"num_instructions": null,
"min_instruction_length": null,
"average_instruction_length": null,
"max_instruction_length": null,
"unique_instructions": null,
"num_top_ranked": 3560,
"min_top_ranked_per_query": 14,
"average_top_ranked_per_query": 353.4244382022472,
"max_top_ranked_per_query": 1245
},
"test_geo": {
"num_samples": 1155697,
"number_of_characters": 371063547,
"num_documents": 1150781,
"min_document_length": 154,
"average_document_length": 315.00053963351843,
"max_document_length": 1289,
"unique_documents": 1150781,
"num_queries": 4916,
"min_query_length": 146,
"average_query_length": 1742.6588689991863,
"max_query_length": 19082,
"unique_queries": 4916,
"none_queries": 0,
"num_relevant_docs": 1150781,
"min_relevant_docs_per_query": 3,
"average_relevant_docs_per_query": 1.0024410089503661,
"max_relevant_docs_per_query": 1274,
"unique_relevant_docs": 1150781,
"num_instructions": null,
"min_instruction_length": null,
"average_instruction_length": null,
"max_instruction_length": null,
"unique_instructions": null,
"num_top_ranked": 4916,
"min_top_ranked_per_query": 3,
"average_top_ranked_per_query": 234.08889340927584,
"max_top_ranked_per_query": 1274
},
"test_vis": {
"num_samples": 1612156,
"number_of_characters": 534911902,
"num_documents": 1606858,
"min_document_length": 154,
"average_document_length": 327.165126601106,
"max_document_length": 1796,
"unique_documents": 1606858,
"num_queries": 5298,
"min_query_length": 176,
"average_query_length": 1737.2595318988297,
"max_query_length": 28468,
"unique_queries": 5298,
"none_queries": 0,
"num_relevant_docs": 1606858,
"min_relevant_docs_per_query": 11,
"average_relevant_docs_per_query": 1.0152887882219706,
"max_relevant_docs_per_query": 1819,
"unique_relevant_docs": 1606858,
"num_instructions": null,
"min_instruction_length": null,
"average_instruction_length": null,
"max_instruction_length": null,
"unique_instructions": null,
"num_top_ranked": 5298,
"min_top_ranked_per_query": 11,
"average_top_ranked_per_query": 303.2952057380143,
"max_top_ranked_per_query": 1819
},
"test_web": {
"num_samples": 837319,
"number_of_characters": 277932894,
"num_documents": 834175,
"min_document_length": 157,
"average_document_length": 326.280188209908,
"max_document_length": 1542,
"unique_documents": 834175,
"num_queries": 3144,
"min_query_length": 146,
"average_query_length": 1831.4624681933842,
"max_query_length": 15329,
"unique_queries": 3144,
"none_queries": 0,
"num_relevant_docs": 834175,
"min_relevant_docs_per_query": 2,
"average_relevant_docs_per_query": 1.0588422391857506,
"max_relevant_docs_per_query": 1064,
"unique_relevant_docs": 834175,
"num_instructions": null,
"min_instruction_length": null,
"average_instruction_length": null,
"max_instruction_length": null,
"unique_instructions": null,
"num_top_ranked": 3144,
"min_top_ranked_per_query": 2,
"average_top_ranked_per_query": 265.3228371501272,
"max_top_ranked_per_query": 1064
}
}
This dataset card was automatically generated using MTEB