mch-dd commited on
Commit
fde10bb
·
verified ·
1 Parent(s): ccb0361

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +91 -3
README.md CHANGED
@@ -1,3 +1,91 @@
1
- ---
2
- license: cc0-1.0
3
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: cc0-1.0
3
+ ---
4
+
5
+ # TED (Tenders Electronic Daily) Dataset Card
6
+
7
+ ## Dataset Overview
8
+
9
+ **Dataset Name:** TED (Tenders Electronic Daily) EU Tenders Dataset
10
+
11
+ **Description:**
12
+ TED (Tenders Electronic Daily) is the online version of the 'Supplement to the Official Journal' of the EU, dedicated to European public procurement. This dataset comprises a collection of procurement notices published by the EU, organized into Parquet files divided by year. Each file contains detailed information about public tenders, including metadata extracted from XML files.
13
+
14
+ **Languages Covered:**
15
+ - German (DEU): 63,128 documents
16
+ - French (FRA): 25,539 documents
17
+ - Polish (POL): 20,594 documents
18
+ - Spanish (SPA): 15,119 documents
19
+ - Dutch (NLD): 12,371 documents
20
+ - Czech (CES): 11,293 documents
21
+ - Romanian (RON): 10,285 documents
22
+ - English (ENG): 10,231 documents
23
+ - Swedish (SWE): 9,560 documents
24
+ - Italian (ITA): 6,603 documents
25
+ - Bulgarian (BUL): 6,281 documents
26
+ - Finnish (FIN): 5,812 documents
27
+ - Latvian (LAV): 5,160 documents
28
+ - Danish (DAN): 3,251 documents
29
+ - Lithuanian (LIT): 3,162 documents
30
+ - Croatian (HR): 3,155 documents
31
+ - Estonian (EST): 2,813 documents
32
+ - Hungarian (HUN): 2,167 documents
33
+ - Portuguese (POR): 2,080 documents
34
+ - Slovenian (SLV): 1,959 documents
35
+ - Slovak (SLK): 1,822 documents
36
+ - Greek (ELL): 1,606 documents
37
+ - Irish (GLE): 1 document
38
+ - Unspecified/Empty: 57 documents
39
+
40
+ **Total Number of Documents:** 224,049
41
+
42
+ **Total Number of Words:** 279,556,839
43
+
44
+ **Average Number of Words per Document:** 1,247.75
45
+
46
+ **Total Number of Characters:** 2,125,102,547
47
+
48
+ **Average Number of Characters per Document:** 9,484.99
49
+
50
+ **Total Number of Rows with Empty Identifier:** 57
51
+
52
+ ## Dataset Structure
53
+
54
+ The dataset is stored in Parquet files organized by year. Each Parquet file contains the following metadata for each document:
55
+
56
+ - `filename`: Name of the original XML file.
57
+ - `identifier`: Unique identifier for the document.
58
+ - `date`: Publication date of the document in DD/MM/YYYY format.
59
+ - `language`: Language of the document.
60
+ - `url`: URL to the document on the TED website.
61
+ - `text`: Full text content of the document.
62
+ - `word_count`: Total number of words in the document.
63
+ - `character_count`: Total number of characters in the document.
64
+
65
+ ## Usage
66
+
67
+ The dataset can be used for various purposes including, but not limited to:
68
+
69
+ - Analyzing public procurement trends across different European countries and languages.
70
+ - Studying the distribution of procurement notices over time.
71
+ - Developing natural language processing models to analyze the content of procurement notices.
72
+ - Extracting insights related to public procurement policies and their implementation across the EU.
73
+
74
+ ## Source
75
+
76
+ The dataset is collected from the Tenders Electronic Daily (TED) platform of the European Union. TED is the official platform for publishing public procurement notices across Europe.
77
+
78
+ ## Dataset Citation
79
+
80
+ If you use this dataset in your research, please cite it as follows:
81
+
82
+ ```
83
+ @dataset{TED_EU_Tenders_2024,
84
+ title={TED (Tenders Electronic Daily) EU Tenders Dataset},
85
+ author={Pleias},
86
+ year={2024},
87
+ description={Collection of EU public procurement notices from the TED platform, organized by year and language, with detailed metadata extracted from XML files.}
88
+ }
89
+ ```
90
+
91
+ **Note:** The dataset is presented and maintained by Pleias. All rights reserved.