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 Translate Test Zh Test

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_translate_test_zh_test

2.59 MB
5137 rows
4 columns


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

Mlqa Translate Train Ar Train

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_translate_train_ar_train

19.3 MB
78058 rows
4 columns


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

Mlqa Translate Train Ar Validation

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_translate_train_ar_validation

2.21 MB
9512 rows
4 columns


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

Mlqa Translate Train De Train

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_translate_train_de_train

17.28 MB
80069 rows
4 columns


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

Mlqa Translate Train De Validation

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_translate_train_de_validation

1.95 MB
9927 rows
4 columns


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

Mlqa Translate Train Es Train

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_translate_train_es_train

17.55 MB
81810 rows
4 columns


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

Mlqa Translate Train Es Validation

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_translate_train_es_validation

1.98 MB
10123 rows
4 columns


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

Mlqa Translate Train Hi Train

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_translate_train_hi_train

26.47 MB
82451 rows
4 columns


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

Mlqa Translate Train Hi Validation

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_translate_train_hi_validation

3.15 MB
10253 rows
4 columns


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

Mlqa Translate Train Vi Train

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_translate_train_vi_train

18.83 MB
84816 rows
4 columns


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

Mlqa Translate Train Vi Validation

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_translate_train_vi_validation

2.15 MB
10356 rows
4 columns


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

Mlqa Translate Train Zh Train

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_translate_train_zh_train

13.65 MB
76285 rows
4 columns


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

Mlqa Translate Train Zh Validation

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_translate_train_zh_validation

1.52 MB
9568 rows
4 columns


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

Mlqa Vi Ar Test

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_vi_ar_test

1.45 MB
2047 rows
4 columns


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

Mlqa Vi Ar Validation

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_vi_ar_validation

122.35 KB
163 rows
4 columns


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

Mlqa Vi De Test

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_vi_de_test

1.07 MB
1675 rows
4 columns


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

Mlqa Vi De Validation

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_vi_de_validation

135.66 KB
182 rows
4 columns


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

Mlqa Vi En Test

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_vi_en_test

3.6 MB
5495 rows
4 columns


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

Mlqa Vi En Validation

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_vi_en_validation

337.07 KB
511 rows
4 columns


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

Mlqa Vi Es Test

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_vi_es_test

1.38 MB
2018 rows
4 columns


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

Mlqa Vi Es Validation

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_vi_es_validation

137.68 KB
189 rows
4 columns


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

Mlqa Vi Hi Test

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_vi_hi_test

1.17 MB
1947 rows
4 columns


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

Mlqa Vi Hi Validation

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_vi_hi_validation

125.92 KB
177 rows
4 columns


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

Mlqa Vi Vi Test

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_vi_vi_test

3.64 MB
5495 rows
4 columns


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

Mlqa Vi Vi Validation

@kaggle.thedevastator_mlqa_multilingual_question_answering_dataset.mlqa_vi_vi_validation

339.95 KB
511 rows
4 columns


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