|
--- |
|
dataset_info: |
|
features: |
|
- name: tokens |
|
sequence: string |
|
- name: ner_tags |
|
sequence: |
|
class_label: |
|
names: |
|
'0': O |
|
'1': B-UoM |
|
'2': I-UoM |
|
'3': B-color |
|
'4': I-color |
|
'5': B-condition |
|
'6': I-condition |
|
'7': B-content |
|
'8': I-content |
|
'9': B-core_product_type |
|
'10': I-core_product_type |
|
'11': B-creator |
|
'12': I-creator |
|
'13': B-department |
|
'14': I-department |
|
'15': B-material |
|
'16': I-material |
|
'17': B-modifier |
|
'18': I-modifier |
|
'19': B-occasion |
|
'20': I-occasion |
|
'21': B-origin |
|
'22': I-origin |
|
'23': B-price |
|
'24': I-price |
|
'25': B-product_name |
|
'26': I-product_name |
|
'27': B-product_number |
|
'28': I-product_number |
|
'29': B-quantity |
|
'30': I-quantity |
|
'31': B-shape |
|
'32': I-shape |
|
'33': B-time |
|
'34': I-time |
|
splits: |
|
- name: train |
|
num_bytes: 553523 |
|
num_examples: 7841 |
|
- name: test |
|
num_bytes: 70308 |
|
num_examples: 993 |
|
- name: validation |
|
num_bytes: 61109 |
|
num_examples: 871 |
|
download_size: 242711 |
|
dataset_size: 684940 |
|
configs: |
|
- config_name: default |
|
data_files: |
|
- split: train |
|
path: data/train-* |
|
- split: test |
|
path: data/test-* |
|
- split: validation |
|
path: data/validation-* |
|
license: cc-by-4.0 |
|
task_categories: |
|
- token-classification |
|
language: |
|
- en |
|
pretty_name: QueryNER |
|
size_categories: |
|
- 1K<n<10K |
|
--- |
|
# Dataset Card for QueryNER |
|
|
|
QueryNER is a sequence labeling dataset for e-commerce query segmentation. |
|
It has 17 different entity types. QueryNER covers nearly the entire query rather than just certain key aspects that may be covered by other aspect-value extraction systems. |
|
|
|
|
|
## Dataset Details |
|
|
|
### Dataset Description |
|
|
|
QueryNER is a manually-annotated dataset and accompanying model for e-commerce query segmentation. Prior work in sequence labeling for e-commerce has largely addressed aspect-value extraction which focuses |
|
on extracting portions of a product title or query for narrowly defined aspects. Our work instead focuses on the goal |
|
of dividing a query into meaningful chunks with broadly applicable types. |
|
QueryNER has 17 different entity types. |
|
|
|
- **Repository:** [QueryNER](https://github.com/bltlab/query-ner) |
|
- **Paper:** Accepted at LREC-COLING 2024, coming soon |
|
|
|
- **Curated by:** BLT Lab |
|
- **Language(s) (NLP):** English |
|
- **License:** CC-BY 4.0 |
|
|
|
### Dataset Sources |
|
|
|
QueryNER is annotation on a subsection of Amazon's [ESCI Shopping Queries dataset](https://github.com/amazon-science/esci-data). |
|
|
|
## Uses |
|
|
|
QueryNER is intended to be used for segmentation of e-commerce queries in English. |
|
|
|
### Direct Use |
|
|
|
QueryNER can be used for research on e-commerce query segmentation. |
|
It may also be used for e-commerce query segmentation for use in further downstream systems; however, we caution users that while the ontology is broadly applicable, using models trained on only this small public release may have suboptimal performance especially on out of domain data. |
|
|
|
### Out-of-Scope Use |
|
|
|
Users would likely experience poor segmentation performance on data outside of the e-commerce domain. |
|
Because the dataset is on the smaller side, additional annotated data on additional data using the QueryNER ontology |
|
may be necessary to get better performance on other datasets. |
|
|
|
|
|
## Dataset Structure |
|
|
|
The dataset includes the query tokens and their tags. |
|
|
|
|
|
## Dataset Creation |
|
See paper. |
|
|
|
### Curation Rationale |
|
|
|
The dataset was created for research and for downstream applications for e-commerce search systems to make use of segmented queries. |
|
|
|
|
|
### Source Data |
|
|
|
The source data is from the Shopping Queries ESCI dataset. |
|
[https://github.com/amazon-science/esci-data](https://github.com/amazon-science/esci-data) |
|
``` |
|
@article{reddy2022shopping, |
|
title={Shopping Queries Dataset: A Large-Scale {ESCI} Benchmark for Improving Product Search}, |
|
author={Chandan K. Reddy and Lluís Màrquez and Fran Valero and Nikhil Rao and Hugo Zaragoza and Sambaran Bandyopadhyay and Arnab Biswas and Anlu Xing and Karthik Subbian}, |
|
year={2022}, |
|
eprint={2206.06588}, |
|
archivePrefix={arXiv} |
|
} |
|
``` |
|
|
|
#### Data Collection and Processing |
|
|
|
See paper |
|
|
|
|
|
#### Who are the source data producers? |
|
|
|
See source data repo and paper. |
|
|
|
|
|
### Annotations |
|
|
|
#### Annotation process |
|
|
|
See paper for details. |
|
|
|
#### Who are the annotators? |
|
|
|
Annotators were contract workers and were paid a living wage. |
|
|
|
#### Personal and Sensitive Information |
|
|
|
The dataset is just user e-commerce queries and should not contain any sensitive information. |
|
|
|
|
|
## Bias, Risks, and Limitations |
|
|
|
The dataset is English only for now. |
|
Bias may be toward e-commerce queries of the source data. |
|
There may also be annotator bias since the dataset is annotated by a single annotator for the training set and three annotators and an adjudicator for the development and test sets. |
|
|
|
|
|
## Citation |
|
|
|
To appear at LREC-COLING 2024. |
|
|
|
**BibTeX:** |
|
|
|
Coming soon |
|
|
|
|
|
|
|
## Dataset Card Authors |
|
|
|
Chester Palen-Michel [@cpalenmichel](https://github.com/cpalenmichel) |
|
|
|
## Dataset Card Contact |
|
|
|
Chester Palen-Michel [@cpalenmichel](https://github.com/cpalenmichel) |