Datasets:

surrey-nlp
/

S3D-v1

Tweet ID int64 1,567,661,715B 1,568,228,015B	Label int64 0 1
1,568,228,015,360,077,800	1
1,568,228,014,990,688,300	1
1,568,228,014,948,761,600	1
1,568,228,014,915,489,800	1
1,568,228,014,617,395,200	1
1,568,228,014,596,440,000	1
1,568,228,014,558,773,200	1
1,568,228,014,550,286,300	1
1,568,228,014,491,574,300	1
1,568,228,014,407,954,400	1
1,568,228,014,340,841,500	1
1,568,228,014,281,867,300	1
1,568,228,014,273,757,200	1
1,568,228,014,210,846,700	1
1,568,228,014,177,284,000	1
1,568,228,014,088,921,000	1
1,568,228,014,085,013,500	1
1,568,228,014,059,569,200	1
1,568,228,014,059,569,200	1
1,568,228,014,009,335,800	1
1,568,228,013,988,257,800	1
1,568,228,013,980,131,300	1
1,568,228,013,938,032,600	1
1,568,228,013,925,359,600	1
1,568,228,013,900,451,800	1
1,568,228,013,896,245,200	1
1,568,228,013,849,862,100	1
1,568,228,013,799,776,300	1
1,568,228,013,699,133,400	1
1,568,228,013,636,034,600	1
1,568,228,013,615,026,200	1
1,568,228,013,593,985,000	1
1,568,228,013,539,737,600	1
1,568,228,013,526,892,500	1
1,568,228,013,510,217,700	1
1,568,228,013,497,634,800	1
1,568,228,013,485,203,500	1
1,568,228,013,485,002,800	1
1,568,228,013,405,249,500	1
1,568,228,013,384,540,200	1
1,568,228,013,329,842,200	1
1,568,228,013,275,512,800	1
1,568,228,013,275,336,700	1
1,568,228,013,258,539,000	1
1,568,228,013,203,984,400	1
1,568,228,013,195,624,400	1
1,568,228,013,182,967,800	1
1,568,228,013,170,479,000	1
1,568,228,013,166,436,400	1
1,568,228,013,153,882,000	1
1,568,228,013,141,106,700	1
1,568,228,013,099,327,500	1
1,568,228,013,099,171,800	1
1,568,228,013,095,137,300	1
1,568,228,013,057,233,000	1
1,568,228,013,040,443,400	1
1,568,228,013,011,275,800	1
1,568,228,013,006,852,000	1
1,568,228,012,977,422,300	1
1,568,228,012,952,326,100	1
1,568,228,012,952,281,000	1
1,568,228,012,939,698,200	1
1,568,228,012,927,365,000	1
1,568,228,012,901,924,900	1
1,568,228,012,868,640,800	1
1,568,228,012,818,301,000	1
1,568,228,012,801,560,600	1
1,568,228,012,801,360,000	1
1,568,228,012,788,957,200	1
1,568,228,012,700,668,000	1
1,568,228,012,696,404,000	1
1,568,228,012,625,199,000	1
1,568,228,012,608,438,300	1
1,568,228,012,587,454,500	1
1,568,228,012,562,301,000	1
1,568,228,012,537,028,600	1
1,568,228,012,532,928,500	1
1,568,228,012,520,243,200	1
1,568,228,012,478,308,400	1
1,568,228,012,470,009,900	1
1,568,228,012,428,267,500	1
1,568,228,012,428,251,100	1
1,568,228,012,407,271,400	1
1,568,228,012,402,819,000	1
1,568,228,012,377,739,300	1
1,568,228,012,356,780,000	1
1,568,228,012,331,618,300	1
1,568,228,012,331,614,200	1
1,568,228,012,319,211,500	1
1,568,228,012,306,440,200	1
1,568,228,012,302,246,000	1
1,568,228,012,302,164,000	1
1,568,228,012,289,818,600	1
1,568,228,012,285,468,700	1
1,568,228,012,184,805,400	1
1,568,228,012,172,124,200	1
1,568,228,012,151,246,800	1
1,568,228,012,138,832,000	1
1,568,228,012,138,586,000	1
1,568,228,012,100,915,200	1

End of preview. Expand in Data Studio

Utilising Weak Supervision to Create S3D: A Sarcasm Annotated Dataset

This is the repository for the S3D dataset published at EMNLP 2022. The dataset can help build sarcasm detection models.

S3D Summary

The S3D dataset is our silver standard dataset of 100,000 tweets labelled for sarcasm using weak supervision by our BERTweet-sarcasm-combined model. These tweets can be accessed by using the Twitter API so that they can be used for other experiments. S3D contains 38879 tweets labelled as sarcastic, and 61211 tweets labelled as not being sarcastic.

Data Fields

Tweet ID: The ID of the labelled tweet
Label: A label to denote if a given tweet is sarcastic

Data Splits

Train: 70,000
Valid: 15,000
Test: 15,000

Downloads last month: 20

Size of downloaded dataset files:

2.2 MB

Size of the auto-converted Parquet files:

843 kB

Number of rows:

100,000