leavoigt commited on
Commit
b706884
·
verified ·
1 Parent(s): 2fa88b6

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +65 -1
README.md CHANGED
@@ -11,4 +11,68 @@ short_description: Discover Development Organization's Projects
11
  license: mit
12
  ---
13
 
14
- Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
11
  license: mit
12
  ---
13
 
14
+ Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
15
+
16
+ Technical Documentation of the system in accordance with EU AI Act.
17
+ System Name: Development Project Synergy Finder
18
+
19
+ Provider / Supplier: GIZ Data Service Center
20
+
21
+ As of: July 2025
22
+
23
+ Multiple international organizations have projects in the same field and region. These projects could collaborate or learn from
24
+ each other to increase their impact if they were aware of one another. The Project Synergy Finder facilitates the search for
25
+ similar projects across different development organizations and banks. Note that this app is a prototype, results may be
26
+ incomplete or inaccurate.
27
+
28
+ ### Models
29
+
30
+ **SDG Text Classifier**
31
+
32
+ Model Name: bert-base-uncased-finetuned-sdg (link will follow soon, if you have questions please contact us at dataservicecenter@giz.de)
33
+ Base Model: [google-bert/bert-base-uncased](https://huggingface.co/google-bert/bert-base-uncased)
34
+ Usage: The SDG categorization in this tool is AI-predicted based on project descriptions and titles using a SDG Classifier trainded on the OSDG dataset.
35
+ License: apache-2.0
36
+ Model Training Data: The training data has been collected by human annotators that are expert in their fields. If you have questions regarding the dataset, please
37
+ reach out to dataservicecenter@giz.de.
38
+ Bias: The data does not contain any known bias, however risk of potential bias can never be fully excluded.
39
+
40
+ ### Data
41
+
42
+ **IATI Data**: The data is sourced from the IATI d-portal, providing project-level information. The International Aid Transparency Initiative (IATI) aims
43
+ to enhance transparency and effectiveness in development cooperation by making data publicly accessible.
44
+ Data Update: The data is updated irregularly, with the last retrieval on 10th May 2024.
45
+ Project Data: Data from projects labeled as active during the last data retrieval are included. The data includes Project Title, Description, URL, Country, and Sector classification (CRS). The CRS5 and CRS3 classifications organize development cooperation into categories, with the 5-digit level providing more specific details within the broader 3-digit categories.
46
+
47
+ **Organizations**: The tool currently includes projects from the following organizations:
48
+
49
+ IAD: Inter-American Development Bank
50
+ ADB: Asian Development Bank
51
+ AfDB: African Development Bank
52
+ EIB: European Investment Bank
53
+ WB: World Bank
54
+ WBTF: World Bank Trust Fund
55
+ BMZ: Federal Ministry for Economic Cooperation and Development (Germany)
56
+ KfW: KfW Development Bank (Germany)
57
+ GIZ: Deutsche Gesellschaft für Internationale Zusammenarbeit (Germany)
58
+ AA: German Federal Foreign Office (Germany)
59
+ Additional Data: The Sustainable Development Goals (SDGs) are 17 UN goals aimed at achieving global sustainability, peace, and prosperity by 2030. The SDG categorization in this tool is AI-predicted based on project descriptions and titles using a SDG Classifier trainded on the OSDG dataset.
60
+
61
+
62
+ ### System Usage
63
+
64
+ The system is a prototype designed to facilitate the quick search of similar projects across different development organizations.
65
+ The system is NOT designed to give a complete overview of all ongoing projects and their matching counterparts. Output may always be incomplete or falsly
66
+ classified and should ALWAYS be reviewed by a human.
67
+ The system does not make autonomous decisions but just provides information.
68
+ No personal data of users is being processed.
69
+ Results are intended for orientation only – not for legal or political advice.
70
+
71
+ ### Transparency Towards Users
72
+ The user interface clearly indicates the use of a fine-tuned text classification transformer models and outlines that the system is a prototype.
73
+
74
+
75
+ ### Contact & Feedback
76
+ Technical development is carried out by the GIZ Data Service Center.
77
+ Please reach out through the contact details provided below, if there are any issues or feedback.
78
+ Contact: For any questions, please contact us via dataservicecenter@giz.de