Name: MLQA - Multilingual Question-Answering
Creator: Kaggle
License: https://creativecommons.org/publicdomain/zero/1.0/

About this Dataset

MLQA - Multilingual Question-Answering

Multilingual Question-Answering Dataset

By mlqa (From Huggingface) [source]

About this dataset

The dataset consists of several files in CSV format that provide context passages or paragraphs along with corresponding questions and answer options. The context passages serve as the source of information from which the questions are derived, and the answer options are potential answers to these questions.

Each file in the dataset contains different language combinations for evaluation purposes. For example, mlqa.es.zh_test.csv focuses on testing multilingual question-answering models in Spanish and Chinese languages. Similarly, mlqa.hi.de_test.csv provides test data specifically for evaluating Hindi-German language pairs.

In order to facilitate accurate evaluation of models' performance, each file includes multiple columns for context and answers. This allows researchers to assess how well their models can generate correct answers based on the given contexts.

Research Ideas

Evaluation of multilingual question-answering models: This dataset can be used to evaluate the performance of different models designed for multilingual question-answering. By providing context, question, and answer pairs in multiple languages, it allows researchers to measure the accuracy and effectiveness of their models across different language pairs.

Cross-lingual transfer learning: The MLQA dataset can be utilized to develop cross-lingual transfer learning techniques. Models trained on this dataset can learn to perform question-answering tasks in one language and then transfer that knowledge to answer questions in another language.

Language understanding research: Researchers studying natural language processing (NLP) and language understanding can use this dataset to analyze how different languages handle questions and answers within various contexts. They can explore linguistic patterns, variations, and differences across languages by comparing the performance of NLP models trained on this dataset for both similar and dissimilar language pairs

Acknowledgements

If you use this dataset in your research, please credit the original authors.
Data Source

License

License: CC0 1.0 Universal (CC0 1.0) - Public Domain Dedication
No Copyright - You can copy, modify, distribute and perform the work, even for commercial purposes, all without asking permission. See Other Information.

Columns

File: mlqa.es.zh_test.csv

Column name	Description
context	The text passage or paragraph in which a question is being asked. (Text)
answers	The possible answers to the question, along with their start and end positions within the context passage. (Text)

File: mlqa.hi.de_test.csv

Column name	Description
context	The text passage or paragraph in which a question is being asked. (Text)
answers	The possible answers to the question, along with their start and end positions within the context passage. (Text)

File: mlqa.zh.de_test.csv

Column name	Description
context	The text passage or paragraph in which a question is being asked. (Text)
answers	The possible answers to the question, along with their start and end positions within the context passage. (Text)

Acknowledgements

If you use this dataset in your research, please credit the original authors.
If you use this dataset in your research, please credit mlqa (From Huggingface).

Tables

Mlqa De Vi Validation

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_de_vi_validation

130.88 KB
182 rows
4 columns


CREATE TABLE mlqa_de_vi_validation (
  "context" VARCHAR,
  "question" VARCHAR,
  "answers" VARCHAR,
  "id" VARCHAR
);

Mlqa De Zh Test

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_de_zh_test

896.08 KB
1621 rows
4 columns


CREATE TABLE mlqa_de_zh_test (
  "context" VARCHAR,
  "question" VARCHAR,
  "answers" VARCHAR,
  "id" VARCHAR
);

Mlqa De Zh Validation

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_de_zh_validation

108.17 KB
190 rows
4 columns


CREATE TABLE mlqa_de_zh_validation (
  "context" VARCHAR,
  "question" VARCHAR,
  "answers" VARCHAR,
  "id" VARCHAR
);

Mlqa En Ar Test

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_en_ar_test

3.42 MB
5335 rows
4 columns


CREATE TABLE mlqa_en_ar_test (
  "context" VARCHAR,
  "question" VARCHAR,
  "answers" VARCHAR,
  "id" VARCHAR
);

Mlqa En Ar Validation

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_en_ar_validation

332.93 KB
517 rows
4 columns


CREATE TABLE mlqa_en_ar_validation (
  "context" VARCHAR,
  "question" VARCHAR,
  "answers" VARCHAR,
  "id" VARCHAR
);

Mlqa En De Test

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_en_de_test

2.74 MB
4517 rows
4 columns


CREATE TABLE mlqa_en_de_test (
  "context" VARCHAR,
  "question" VARCHAR,
  "answers" VARCHAR,
  "id" VARCHAR
);

Mlqa En De Validation

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_en_de_validation

325.3 KB
512 rows
4 columns


CREATE TABLE mlqa_en_de_validation (
  "context" VARCHAR,
  "question" VARCHAR,
  "answers" VARCHAR,
  "id" VARCHAR
);

Mlqa En En Test

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_en_en_test

7.21 MB
11590 rows
4 columns


CREATE TABLE mlqa_en_en_test (
  "context" VARCHAR,
  "question" VARCHAR,
  "answers" VARCHAR,
  "id" VARCHAR
);

Mlqa En En Validation

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_en_en_validation

695.04 KB
1148 rows
4 columns


CREATE TABLE mlqa_en_en_validation (
  "context" VARCHAR,
  "question" VARCHAR,
  "answers" VARCHAR,
  "id" VARCHAR
);

Mlqa En Es Test

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_en_es_test

3.29 MB
5253 rows
4 columns


CREATE TABLE mlqa_en_es_test (
  "context" VARCHAR,
  "question" VARCHAR,
  "answers" VARCHAR,
  "id" VARCHAR
);

Mlqa En Es Validation

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_en_es_validation

309.29 KB
500 rows
4 columns


CREATE TABLE mlqa_en_es_validation (
  "context" VARCHAR,
  "question" VARCHAR,
  "answers" VARCHAR,
  "id" VARCHAR
);

Mlqa En Hi Test

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_en_hi_test

3.07 MB
4918 rows
4 columns


CREATE TABLE mlqa_en_hi_test (
  "context" VARCHAR,
  "question" VARCHAR,
  "answers" VARCHAR,
  "id" VARCHAR
);

Mlqa En Hi Validation

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_en_hi_validation

314.15 KB
507 rows
4 columns


CREATE TABLE mlqa_en_hi_validation (
  "context" VARCHAR,
  "question" VARCHAR,
  "answers" VARCHAR,
  "id" VARCHAR
);

Mlqa En Vi Test

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_en_vi_test

3.65 MB
5495 rows
4 columns


CREATE TABLE mlqa_en_vi_test (
  "context" VARCHAR,
  "question" VARCHAR,
  "answers" VARCHAR,
  "id" VARCHAR
);

Mlqa En Vi Validation

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_en_vi_validation

345.16 KB
511 rows
4 columns


CREATE TABLE mlqa_en_vi_validation (
  "context" VARCHAR,
  "question" VARCHAR,
  "answers" VARCHAR,
  "id" VARCHAR
);

Mlqa En Zh Test

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_en_zh_test

3.45 MB
5137 rows
4 columns


CREATE TABLE mlqa_en_zh_test (
  "context" VARCHAR,
  "question" VARCHAR,
  "answers" VARCHAR,
  "id" VARCHAR
);

Mlqa En Zh Validation

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_en_zh_validation

324.84 KB
504 rows
4 columns


CREATE TABLE mlqa_en_zh_validation (
  "context" VARCHAR,
  "question" VARCHAR,
  "answers" VARCHAR,
  "id" VARCHAR
);

Mlqa Es Ar Test

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_es_ar_test

981.24 KB
1978 rows
4 columns


CREATE TABLE mlqa_es_ar_test (
  "context" VARCHAR,
  "question" VARCHAR,
  "answers" VARCHAR,
  "id" VARCHAR
);

Mlqa Es Ar Validation

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_es_ar_validation

87.5 KB
161 rows
4 columns


CREATE TABLE mlqa_es_ar_validation (
  "context" VARCHAR,
  "question" VARCHAR,
  "answers" VARCHAR,
  "id" VARCHAR
);

Mlqa Es De Test

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_es_de_test

814.17 KB
1776 rows
4 columns


CREATE TABLE mlqa_es_de_test (
  "context" VARCHAR,
  "question" VARCHAR,
  "answers" VARCHAR,
  "id" VARCHAR
);

Mlqa Es De Validation

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_es_de_validation

95.68 KB
196 rows
4 columns


CREATE TABLE mlqa_es_de_validation (
  "context" VARCHAR,
  "question" VARCHAR,
  "answers" VARCHAR,
  "id" VARCHAR
);

Mlqa Es En Test

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_es_en_test

2.45 MB
5253 rows
4 columns


CREATE TABLE mlqa_es_en_test (
  "context" VARCHAR,
  "question" VARCHAR,
  "answers" VARCHAR,
  "id" VARCHAR
);

Mlqa Es En Validation

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_es_en_validation

248.4 KB
500 rows
4 columns


CREATE TABLE mlqa_es_en_validation (
  "context" VARCHAR,
  "question" VARCHAR,
  "answers" VARCHAR,
  "id" VARCHAR
);

Mlqa Es Es Test

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_es_es_test

2.46 MB
5253 rows
4 columns


CREATE TABLE mlqa_es_es_test (
  "context" VARCHAR,
  "question" VARCHAR,
  "answers" VARCHAR,
  "id" VARCHAR
);

Mlqa Es Es Validation

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_es_es_validation

250 KB
500 rows
4 columns


CREATE TABLE mlqa_es_es_validation (
  "context" VARCHAR,
  "question" VARCHAR,
  "answers" VARCHAR,
  "id" VARCHAR
);