File size: 1,855 Bytes
07b40bd
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
---
license: apache-2.0
---
Medium Article Dataset For Extractive Text Summarization
Author
Vaibhav Gulati, Deepika Kumar
(Department of Computer Science and Engineering, Bharati Vidyapeeth’s College of Engineering, New Delhi 110063, India)

Contact
Contact using the following email: gulvaibhav20@gmail.com

Subject
Engineering

Description
The dataset contains 75 articles that belongs to a variety of different categories including technology, science, business, nature and entertainment etc. These articles were published on the Medium website in 2018 and were extracted using web scraping tools and techniques. The dataset contains the textual content of the articles along with information about the publication date, sub-title, author, word count, etc. In addition to this, the dataset contains human generated summaries of all the 75 articles, where the length of each summary is about one-fifth of the length of the original Medium article.

Dataset Parameters	Values
Number of Articles (Instances)	75
Categories	25
Average number of sentences per article	89.5
Maximum number of sentences per article	229
Minimum number of sentences per article	30
File Description
Dataset.csv - This file contains the whole Medium article dataset having the attributes:

Attribute	Information
Title	Title of the article
Sub_Title	Sub-title of the article
Text	The textual content of the article
Word_Count	Number of Words present in article
Category	Genre of the article
Author	Author of the article
URL	Medium link for the article
Human Generated Summary	Human generated extractive summary of the article
Dataset Usage
The dataset can be used for the evaluation of the Extractive Text Summarization process. The summary of the articles generated by the extractive text summarizer model can be compared with the Human Generated Summary provided in the dataset.