MohammadOthman commited on
Commit
e7372c4
1 Parent(s): c9088e2

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +167 -1
README.md CHANGED
@@ -1,3 +1,169 @@
1
  ---
2
- license: cc-by-nc-4.0
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3
  ---
 
 
 
1
  ---
2
+ tags:
3
+ - text classification
4
+ - clustering
5
+ - newsgroups
6
+ license: unknown
7
+ size: 70 MB
8
+ language:
9
+ - en
10
+ description: >
11
+ The 20 Newsgroups dataset is a collection of approximately 20,000 newsgroup
12
+ documents, partitioned across 20 different newsgroups. It's widely used for
13
+ text classification and clustering experiments. The dataset offers three
14
+ versions: the original, a date-sorted version, and a version with only "From"
15
+ and "Subject" headers.
16
+ homepage: http://qwone.com/~jason/20Newsgroups/
17
+ task_categories:
18
+ - text-classification
19
+ - summarization
20
+ - question-answering
21
+ ---
22
+
23
+ # 20 Newsgroups Dataset
24
+
25
+ ## Introduction
26
+
27
+ The 20 Newsgroups dataset comprises roughly 20,000 documents from newsgroups, with an almost even distribution across 20 distinct newsgroups. Initially gathered by Ken Lang, this dataset has gained prominence in the machine learning community, particularly for text-related applications like classification and clustering.
28
+
29
+ ## Dataset Structure
30
+
31
+ The dataset's organization is based on 20 different newsgroups, each representing a unique topic. While some of these newsgroups share similarities or are closely related, others are quite distinct from one another.
32
+
33
+ ### List of Newsgroups:
34
+
35
+ - Computer Graphics
36
+ - Windows OS Miscellaneous
37
+ - IBM PC Hardware
38
+ - Mac Hardware
39
+ - Windows X
40
+ - Automobiles
41
+ - Motorcycles
42
+ - Baseball
43
+ - Hockey
44
+ - Cryptography
45
+ - Electronics
46
+ - Medicine
47
+ - Space
48
+ - Miscellaneous Sales
49
+ - Miscellaneous Politics
50
+ - Politics & Guns
51
+ - Middle East Politics
52
+ - Miscellaneous Religion
53
+ - Atheism
54
+ - Christianity
55
+
56
+ ## Sample Entries
57
+
58
+ ### Sample from `Windows X`
59
+
60
+ ```
61
+ From: Bill.Kayser@delft.SGp.slb.COM (Bill Kayser)
62
+ Subject: Re: TeleUse, UIM/X, and C++
63
+ Article-I.D.: parsival.199304060629.AA00339
64
+ Organization: The Internet
65
+ Lines: 25
66
+ NNTP-Posting-Host: enterpoop.mit.edu
67
+ To: xpert@expo.lcs.mit.edu
68
+ Cc: Bill.Kayser@delft.sgp.slb.com
69
+
70
+
71
+ >
72
+ > Does anyone have any good ideas on how to integrate C++ code elegantly
73
+ > with TeleUse, UIM/X / Interface Architect generated code?
74
+ >
75
+ > Source would be great, but any suggestions are welcome.
76
+
77
+ It's my understanding that the next release of UIM/X, due out
78
+ last February :-) has full support for C++.
79
+
80
+ I use XDesigner which does not have the interpreter or UI meta languages
81
+ of these other tools but does fully support C++ code generation,
82
+ reusable templates via C++ classes which are generated, a variety of
83
+ other handy features for using C++ and layout functions in different
84
+ ways, and generates Motif 1.2 code (including drag 'n drop,
85
+ internationalization, etc.). Fits in quite nicely with Doug Young's
86
+ paradigm for C++/Motif.
87
+
88
+ Available in the US from VI Corp, in Europe from Imperial Software,
89
+ London (see FAQ for details).
90
+
91
+ Bill
92
+ ________________________________________________________________________
93
+ Schlumberger Geco Prakla
94
+ kayser@delft.sgp.slb.com
95
+
96
+ ```
97
+
98
+
99
+ ### Sample from `Electronics`
100
+
101
+ ```
102
+ From: baden@sys6626.bison.mb.ca (baden de bari)
103
+ Subject: Re: Jacob's Ladder
104
+ Organization: System 6626 BBS, Winnipeg Manitoba Canada
105
+ Lines: 36
106
+
107
+ g92m3062@alpha.ru.ac.za (Brad Meier) writes:
108
+
109
+ > Hi, I'm looking for a circuit, that is called a "Jacob's Ladder".
110
+ > This little box is usually seen in sci-fi movies. It consists of
111
+ > two curves of wire protruding into the air, with little blue sparks
112
+ > starting at their base (where the two wires are closer to each other),
113
+ > moving up the wires to the top, and ending in a small crackling noise.
114
+ >
115
+ > Could anyone supply me with the schematic for the innards of this box?
116
+ >
117
+ > Thanks in advance
118
+ > Mike
119
+ >
120
+ > (Please reply by email to g90k3853@alpha.ru.ac.za)
121
+ >
122
+ > --
123
+ > | / | | ~|~ /~~\ | | ~|~ /~~\ |~~\ /~~\ The KnightOrc
124
+ > |/ |\ | | | __ |__| | | | |__/ | g92m3062@hippo.ru.ac.za
125
+ > |\ | \| | | | | | | | | | | | "When it's over I'll go home,
126
+ > | \ | | _|_ \__/ | | | \__/ | | \__/ until then, I stay!" - Me
127
+
128
+
129
+ I'd like any accumulated information on this as well please.
130
+
131
+ Thanks.
132
+
133
+ _________________________________________
134
+ _____ |
135
+ | | | |
136
+ =========== | Baden de Bari |
137
+ | o o | | |
138
+ | ^ | | baden@sys6626.bison.ca |
139
+ | {-} | | baden@inqmind.bison.ca |
140
+ \_____/ | |
141
+ -----------------------------------------
142
+
143
+ ```
144
+
145
+ ## Data Availability
146
+
147
+ The dataset is bundled in `.tar.gz` format. Within each bundle, individual subdirectories represent a newsgroup. Every file within these subdirectories corresponds to a document posted in that specific newsgroup.
148
+
149
+ There are three primary versions of the dataset:
150
+
151
+ 1. The original version, which remains unaltered.
152
+ 2. A version sorted by date, which segregates the data into training (60%) and test (40%) sets. This version has removed duplicates and some headers for clarity.
153
+ 3. A version that only retains the "From" and "Subject" headers, with duplicates removed.
154
+
155
+ For those seeking a more consistent benchmark, the date-sorted version is recommended. It offers a realistic split based on time and has removed any newsgroup-specific identifiers.
156
+
157
+ ## Matlab/Octave Version
158
+
159
+ For users of Matlab or Octave, a processed variant of the date-sorted dataset is available. This version is structured as a sparse matrix and includes files like `train.data`, `train.label`, `test.data`, and more. Additionally, a vocabulary file is provided to help users understand the indexed data.
160
+
161
+ ## Additional Information
162
+
163
+ For more details and the original dataset, you can refer to the [official website](http://qwone.com/~jason/20Newsgroups/).
164
+
165
+
166
+
167
  ---
168
+ license: cc-by-nc-4.0
169
+ ---