Datasets:

Modalities:
Text
Languages:
English
ArXiv:
Libraries:
Datasets
License:
nbroad HF staff commited on
Commit
42aac7a
1 Parent(s): 0f2e4bd

add readme

Browse files
Files changed (1) hide show
  1. README.md +92 -0
README.md ADDED
@@ -0,0 +1,92 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ languages:
3
+ - en
4
+ multilinguality:
5
+ - monolingual
6
+ size_categories:
7
+ - 100K<n<1M
8
+ task_categories:
9
+ - summarization
10
+ ---
11
+ # MediaSum
12
+ ## Description
13
+ This large-scale media interview dataset contains 463.6K transcripts with abstractive summaries,
14
+ collected from interview transcripts and overview / topic descriptions from NPR and CNN.
15
+
16
+ ### **NOTE: The authors have requested that this dataset be used for research purposes only**
17
+
18
+
19
+ ## Homepage
20
+ https://github.com/zcgzcgzcg1/MediaSum
21
+ ## Paper
22
+ https://arxiv.org/abs/2103.06410
23
+ ## Authors
24
+ ### Chenguang Zhu*, Yang Liu*, Jie Mei, Michael Zeng
25
+ #### Microsoft Cognitive Services Research Group
26
+ {chezhu,yaliu10,jimei,nzeng}@microsoft.com
27
+
28
+ ## Citation
29
+
30
+ @article{zhu2021mediasum,
31
+ title={MediaSum: A Large-scale Media Interview Dataset for Dialogue Summarization},
32
+ author={Zhu, Chenguang and Liu, Yang and Mei, Jie and Zeng, Michael},
33
+ journal={arXiv preprint arXiv:2103.06410},
34
+ year={2021}
35
+ }
36
+
37
+ ## Dataset size
38
+ Train: 443,596
39
+ Validation: 10,000
40
+ Test: 10,000
41
+ ## Data details
42
+ - id (string): unique identifier
43
+ - program (string): the program this transcript came from
44
+ - date (string): date of program
45
+ - url (string): link to where audio and transcript are located
46
+ - title (string): title of the program
47
+ - summary (string): summary of the program
48
+ - utt (list): list of utterances by the speakers in the program. corresponds with `speaker`
49
+ - speaker (list): list of speakers, corresponds with `utt`
50
+
51
+
52
+ Example:
53
+ ```
54
+ {
55
+ "id": "NPR-11",
56
+ "program": "Day to Day",
57
+ "date": "2008-06-10",
58
+ "url": "https://www.npr.org/templates/story/story.php?storyId=91356794",
59
+ "title": "Researchers Find Discriminating Plants",
60
+ "summary": "The \"sea rocket\" shows preferential treatment to plants that are its kin. Evolutionary plant ecologist Susan Dudley of McMaster University in Ontario discusses her discovery.",
61
+ "utt": [
62
+ "This is Day to Day. I'm Madeleine Brand.",
63
+ "And I'm Alex Cohen.",
64
+ "Coming up, the question of who wrote a famous religious poem turns into a very unchristian battle.",
65
+ "First, remember the 1970s? People talked to their houseplants, played them classical music. They were convinced plants were sensuous beings and there was that 1979 movie, \"The Secret Life of Plants.\"",
66
+ "Only a few daring individuals, from the scientific establishment, have come forward with offers to replicate his experiments, or test his results. The great majority are content simply to condemn his efforts without taking the trouble to investigate their validity.",
67
+ ...
68
+ "OK. Thank you.",
69
+ "That's Susan Dudley. She's an associate professor of biology at McMaster University in Hamilt on Ontario. She discovered that there is a social life of plants."
70
+ ],
71
+ "speaker": [
72
+ "MADELEINE BRAND, host",
73
+ "ALEX COHEN, host",
74
+ "ALEX COHEN, host",
75
+ "MADELEINE BRAND, host",
76
+ "Unidentified Male",
77
+ ..."
78
+ Professor SUSAN DUDLEY (Biology, McMaster University)",
79
+ "MADELEINE BRAND, host"
80
+ ]
81
+ }
82
+ ```
83
+ ## Using the dataset
84
+ ```python
85
+ from datasets import load_dataset
86
+ ds = load_dataset("nbroad/mediasum")
87
+ ```
88
+ ## Data location
89
+ https://drive.google.com/file/d/1ZAKZM1cGhEw2A4_n4bGGMYyF8iPjLZni/view?usp=sharing
90
+
91
+ ## License
92
+ No license specified, but the authors have requested that this dataset be used for research purposes only.