freinold commited on
Commit
01073f8
·
1 Parent(s): bfbf4eb

Finished english dataset card and added german dataset card.

Browse files
Files changed (2) hide show
  1. README.md +53 -63
  2. README_de.md +135 -0
README.md CHANGED
@@ -49,7 +49,9 @@ pretty_name: Munich Public Services / Münchner Verwaltungs-Dienstleistungen
49
  size_categories:
50
  - 1K<n<10K
51
  ---
52
- # Dataset Card for Munich Public Services (🇩🇪 below)
 
 
53
 
54
  <!-- Provide a quick summary of the dataset. -->
55
  This dataset contains information about the services provided to the public by the City of Munich in the form of written articles, corresponding metadata as well as embeddings.
@@ -64,18 +66,14 @@ Next to the content of the articles, the dataset also includes metadata such the
64
  The dataset also includes embedding vectors computed from each article, which can be used for Retrieval or other NLP tasks.
65
 
66
  - **Curated by:** City of Munich, it@M, KI Competence Center
67
- - **Language(s) (NLP):** German, English
68
  - **License:** MIT
69
 
70
  ### Dataset Sources
71
 
72
  <!-- Provide the basic links for the dataset. -->
73
- The dataset is sourced from the city of Munich's public services website and it's underlying CMS.
74
- All articles were written by the city of Munich's employees.
75
- The other metadata was extracted from the CMS.
76
- The embeddings were generated using OPENAI's `text-embedding-3-large` model.
77
-
78
- - **Homepage:** [Munich Public Services](https://www.muenchen.de/rathaus/home.html)
79
 
80
  ## Uses
81
 
@@ -84,105 +82,97 @@ The embeddings were generated using OPENAI's `text-embedding-3-large` model.
84
  ### Direct Use
85
 
86
  <!-- This section describes suitable use cases for the dataset. -->
87
-
88
- [More Information Needed]
 
89
 
90
  ### Out-of-Scope Use
91
 
92
  <!-- This section addresses misuse, malicious use, and uses that the dataset will not work well for. -->
 
 
93
 
94
- [More Information Needed]
 
95
 
96
  ## Dataset Structure
97
 
98
  <!-- This section provides a description of the dataset fields, and additional information about the dataset structure such as criteria used to create the splits, relationships between data points, etc. -->
99
-
100
- [More Information Needed]
 
 
 
 
 
 
 
 
 
 
 
 
101
 
102
  ## Dataset Creation
103
 
104
  ### Curation Rationale
105
 
106
  <!-- Motivation for the creation of this dataset. -->
107
-
108
- [More Information Needed]
109
 
110
  ### Source Data
111
 
112
  <!-- This section describes the source data (e.g. news text and headlines, social media posts, translated sentences, ...). -->
 
 
113
 
114
  #### Data Collection and Processing
115
 
116
  <!-- This section describes the data collection and processing process such as data selection criteria, filtering and normalization methods, tools and libraries used, etc. -->
117
-
118
- [More Information Needed]
 
 
119
 
120
  #### Who are the source data producers?
121
 
122
  <!-- This section describes the people or systems who originally created the data. It should also include self-reported demographic or identity information for the source data creators if this information is available. -->
123
-
124
- [More Information Needed]
125
-
126
- ### Annotations [optional]
127
-
128
- <!-- If the dataset contains annotations which are not part of the initial data collection, use this section to describe them. -->
129
-
130
- #### Annotation process
131
-
132
- <!-- This section describes the annotation process such as annotation tools used in the process, the amount of data annotated, annotation guidelines provided to the annotators, interannotator statistics, annotation validation, etc. -->
133
-
134
- [More Information Needed]
135
-
136
- #### Who are the annotators?
137
-
138
- <!-- This section describes the people or systems who created the annotations. -->
139
-
140
- [More Information Needed]
141
 
142
  #### Personal and Sensitive Information
143
 
144
  <!-- State whether the dataset contains data that might be considered personal, sensitive, or private (e.g., data that reveals addresses, uniquely identifiable names or aliases, racial or ethnic origins, sexual orientations, religious beliefs, political opinions, financial or health data, etc.). If efforts were made to anonymize the data, describe the anonymization process. -->
145
-
146
- [More Information Needed]
147
 
148
  ## Bias, Risks, and Limitations
149
 
150
  <!-- This section is meant to convey both technical and sociotechnical limitations. -->
 
 
151
 
152
- [More Information Needed]
 
 
 
 
 
153
 
154
  ### Recommendations
155
 
156
  <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
 
 
157
 
158
- Users should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.
159
-
160
- ## Citation [optional]
161
-
162
- <!-- If there is a paper or blog post introducing the dataset, the APA and Bibtex information for that should go in this section. -->
163
-
164
- **BibTeX:**
165
-
166
- [More Information Needed]
167
-
168
- **APA:**
169
-
170
- [More Information Needed]
171
-
172
- ## Glossary [optional]
173
-
174
- <!-- If relevant, include terms and calculations in this section that can help readers understand the dataset or dataset card. -->
175
-
176
- [More Information Needed]
177
-
178
- ## More Information [optional]
179
 
180
- [More Information Needed]
181
 
182
- ## Dataset Card Authors [optional]
183
 
184
- [More Information Needed]
185
 
186
  ## Dataset Card Contact
187
 
188
- [More Information Needed]
 
49
  size_categories:
50
  - 1K<n<10K
51
  ---
52
+ # Dataset Card for "Munich Public Services"
53
+
54
+ [**zur deutschen Version wechseln**](README_de.md)
55
 
56
  <!-- Provide a quick summary of the dataset. -->
57
  This dataset contains information about the services provided to the public by the City of Munich in the form of written articles, corresponding metadata as well as embeddings.
 
66
  The dataset also includes embedding vectors computed from each article, which can be used for Retrieval or other NLP tasks.
67
 
68
  - **Curated by:** City of Munich, it@M, KI Competence Center
69
+ - **Language(s):** German, English
70
  - **License:** MIT
71
 
72
  ### Dataset Sources
73
 
74
  <!-- Provide the basic links for the dataset. -->
75
+ - **City of Munich AI Documentation:** [ki.muenchen.de](https://ki.muenchen.de/)
76
+ - **Dataset Source Webpages:** [Munich Public Services](https://stadt.muenchen.de/service/en-GB/)
 
 
 
 
77
 
78
  ## Uses
79
 
 
82
  ### Direct Use
83
 
84
  <!-- This section describes suitable use cases for the dataset. -->
85
+ The dataset can be used to provide information about the public services provided by the City of Munich.
86
+ This is especially useful for grounding information retrieval systems or Large Language Models in the context of public services.
87
+ The data can also be used to analyze itself for patterns or trends in the public services provided by the City of Munich.
88
 
89
  ### Out-of-Scope Use
90
 
91
  <!-- This section addresses misuse, malicious use, and uses that the dataset will not work well for. -->
92
+ The dataset should not be used for any purpose other than the intended use case of providing information about the public services provided by the City of Munich.
93
+ This also includes other cities or government entities, as the information is specific to Munich.
94
 
95
+ The dataset should not be used for any malicious purposes like desinformation or fake news generation.
96
+ Also using the dataset as input to an LLM for generation of more 'fake' articles should not be done.
97
 
98
  ## Dataset Structure
99
 
100
  <!-- This section provides a description of the dataset fields, and additional information about the dataset structure such as criteria used to create the splits, relationships between data points, etc. -->
101
+ The dataset consists of the following fields:
102
+
103
+ | Field | Type | Description |
104
+ |-------------------|------------|-----------------------------------------------------------------------------------------------|
105
+ | `id` | int | The unique identifier of the article. |
106
+ | `name` | string | The title of the article. |
107
+ | `content` | string | The full text content of the article, formatted in Markdown. |
108
+ | `language` | string | The language of the article, either German (`de`) or English (`en`). |
109
+ | `description` | string | A short description of the article content with a maximum length of x characters, formatted in HTML. |
110
+ | `source` | string | Link to the source webpage of the article. |
111
+ | `public` | bool | Boolean value indicating if the article is public or not, always `True`. |
112
+ | `lastModification`| timestamp | Timestamp of the last modification of the article in UTC. |
113
+ | `keywords` | sequence [string] | A list of keywords associated with the article, the keywords might be in German or English. |
114
+ | `embedding` | sequence [float] | A 3072 dimensional embedding vector computed from the article content, using OpenAI's [`text-embedding-3-large`](https://platform.openai.com/docs/guides/embeddings/#embedding-models) model. |
115
 
116
  ## Dataset Creation
117
 
118
  ### Curation Rationale
119
 
120
  <!-- Motivation for the creation of this dataset. -->
121
+ The dataset was created as a side product of a project to provide a better information retrieval system for the City of Munich's public services.
122
+ Periodically the dataset is updated to include new articles and to update the embeddings.
123
 
124
  ### Source Data
125
 
126
  <!-- This section describes the source data (e.g. news text and headlines, social media posts, translated sentences, ...). -->
127
+ The dataset is sourced from the city of Munich's public services website and it's underlying CMS.
128
+ All articles were written by the city of Munich's employees.
129
 
130
  #### Data Collection and Processing
131
 
132
  <!-- This section describes the data collection and processing process such as data selection criteria, filtering and normalization methods, tools and libraries used, etc. -->
133
+ The data was collected by querying an internal API of the city of Munich's CMS.
134
+ The data was then processed to extract the metadata and the content of the articles.
135
+ The article content was transformed from HTML to Markdown by using the [markdownify](https://pypi.org/project/markdownify/) library.
136
+ The embeddings were generated using OpenAI's [`text-embedding-3-large`](https://platform.openai.com/docs/guides/embeddings/#embedding-models) model.
137
 
138
  #### Who are the source data producers?
139
 
140
  <!-- This section describes the people or systems who originally created the data. It should also include self-reported demographic or identity information for the source data creators if this information is available. -->
141
+ The source data producers are the employees of the City of Munich who wrote the articles, there is no attribution to individual authors.
142
+ The employees are a diverse group of people working in the city's administration with special knowledge about the public services of their respective departments and skills in writing understandable articles.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
143
 
144
  #### Personal and Sensitive Information
145
 
146
  <!-- State whether the dataset contains data that might be considered personal, sensitive, or private (e.g., data that reveals addresses, uniquely identifiable names or aliases, racial or ethnic origins, sexual orientations, religious beliefs, political opinions, financial or health data, etc.). If efforts were made to anonymize the data, describe the anonymization process. -->
147
+ The dataset does not contain any personal or sensitive information.
 
148
 
149
  ## Bias, Risks, and Limitations
150
 
151
  <!-- This section is meant to convey both technical and sociotechnical limitations. -->
152
+ The provided dataset is only a subset of all public services provided by the City of Munich.
153
+ Besides that, the data can contain biases in the form of
154
 
155
+ - the language used in the articles (so called "Beamtendeutsch" / "administrative German")
156
+ - the quality and information richness of the articles
157
+ - the number of articles for one topic
158
+ - and the keywords associated with the articles.
159
+
160
+ This can lead to biases in the embeddings and the information retrieval system built on top of the dataset.
161
 
162
  ### Recommendations
163
 
164
  <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
165
+ To mitigate the biases in the dataset, it is recommended to use the dataset in combination with other sources of information.
166
+ Another approach is to evaluate the dataset with a custom evaluation set to identify biases for the specific use case and to mitigate them.
167
 
168
+ ## More Information
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
169
 
170
+ For more information about the AI projects of the City of Munich, please visit the [ki.muenchen.de](https://ki.muenchen.de/) website.
171
 
172
+ ## Dataset Card Authors
173
 
174
+ Fabian Reinold, City of Munich, it@M, AI Competence Center
175
 
176
  ## Dataset Card Contact
177
 
178
+ For any questions about the dataset, please [contact us](mailto:[email protected])
README_de.md ADDED
@@ -0,0 +1,135 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Dataset Card für "Münchner Verwaltungs-Dienstleistungen"
2
+
3
+ [**switch to English version**](README.md)
4
+
5
+ <!-- Provide a quick summary of the dataset. -->
6
+ Dieser Datensatz enthält Informationen über die von der Landeshauptstadt München angebotenen Dienstleistungen in Form von schriftlichen Artikeln, zugehörigen Metadaten sowie Einbettungen.
7
+
8
+ ## Datensatzdetails
9
+
10
+ ### Datensatzbeschreibung
11
+
12
+ <!-- Provide a longer summary of what this dataset is. -->
13
+ Der Datensatz "Münchner Verwaltungs-Dienstleistungen" enthält etwa 1.400 Artikel über verschiedene öffentliche Dienstleistungen der Landeshauptstadt München.
14
+ Neben dem Inhalt der Artikel enthält der Datensatz auch Metadaten wie die Sprache des Artikels, das Datum der letzten Änderung und die mit dem Artikel verbundenen Schlüsselwörter.
15
+ Der Datensatz enthält auch Einbettungsvektoren, die aus jedem Artikel berechnet wurden und für Retrieval- oder andere NLP-Aufgaben verwendet werden können.
16
+
17
+ - **Kuratiert von:** Landeshauptstadt München, it@M, KI Competence Center
18
+ - **Sprache(n) (NLP):** Deutsch, Englisch
19
+ - **Lizenz:** MIT
20
+
21
+ ### Datenquellen
22
+
23
+ <!-- Provide the basic links for the dataset. -->
24
+ - **KI-Dokumentation der Landeshauptstadt München:** [ki.muenchen.de](https://ki.muenchen.de/)
25
+ - **Datensatz-Quellwebseiten:** [Münchner Verwaltungs-Dienstleistungen](https://stadt.muenchen.de/service/en-GB/)
26
+
27
+ ## Verwendungszwecke
28
+
29
+ <!-- Address questions around how the dataset is intended to be used. -->
30
+
31
+ ### Direkte Nutzung
32
+
33
+ <!-- This section describes suitable use cases for the dataset. -->
34
+ Der Datensatz kann verwendet werden, um Informationen über die öffentlichen Dienstleistungen der Landeshauptstadt München bereitzustellen.
35
+ Dies ist besonders nützlich für KI-basierte Suchsysteme oder das Anreichern großer Sprachmodelle im Kontext öffentlicher Dienstleistungen.
36
+ Die Daten können auch verwendet werden, um Muster oder Trends in den von der Landeshauptstadt München angebotenen öffentlichen Dienstleistungen zu analysieren.
37
+
38
+ ### Nicht geeignete Nutzung
39
+
40
+ <!-- This section addresses misuse, malicious use, and uses that the dataset will not work well for. -->
41
+ Der Datensatz sollte nicht für andere Zwecke als den vorgesehenen Verwendungszweck der Bereitstellung von Informationen über die öffentlichen Dienstleistungen der Landeshauptstadt München verwendet werden.
42
+ Dies gilt auch für andere Städte oder Regierungsstellen, da die Informationen spezifisch für München sind.
43
+
44
+ Der Datensatz sollte nicht für böswillige Zwecke wie Desinformation oder die Erstellung von Fake News verwendet werden.
45
+ Auch die Verwendung des Datensatzes als Eingabe für ein LLM zur Erstellung weiterer "Fake"-Artikel sollte unterlassen werden.
46
+
47
+ ## Datenstruktur
48
+
49
+ <!-- This section provides a description of the dataset fields, and additional information about the dataset structure such as criteria used to create the splits, relationships between data points, etc. -->
50
+ Der Datensatz besteht aus den folgenden Feldern:
51
+
52
+ | Feld | Typ | Beschreibung |
53
+ |-------------------|------------|-----------------------------------------------------------------------------------------------|
54
+ | `id` | int | Die eindeutige ID des Artikels. |
55
+ | `name` | string | Der Titel des Artikels. |
56
+ | `content` | string | Der vollständige Textinhalt des Artikels, formatiert in Markdown. |
57
+ | `language` | string | Die Sprache des Artikels, entweder Deutsch (`de`) oder Englisch (`en`). |
58
+ | `description` | string | Eine kurze Beschreibung des Artikelinhalts mit einer maximalen Länge von x Zeichen, formatiert in HTML. |
59
+ | `source` | string | Link zur Quellwebseite des Artikels. |
60
+ | `public` | bool | Wahrheitswert, der angibt, ob der Artikel öffentlich ist oder nicht, immer `True`. |
61
+ | `lastModification`| timestamp | Zeitstempel der letzten Änderung des Artikels in UTC. |
62
+ | `keywords` | sequence [string] | Eine Liste von Schlüsselwörtern, die mit dem Artikel verbunden sind; die Schlüsselwörter können auf Deutsch oder Englisch sein. |
63
+ | `embedding` | sequence [float] | Ein 3072-dimensionaler Einbettungsvektor, der aus dem Artikelinhalt berechnet wurde, unter Verwendung von OpenAI's [`text-embedding-3-large`](https://platform.openai.com/docs/guides/embeddings/#embedding-models) KI-Modell. |
64
+
65
+ ## Datensatz-Erstellung
66
+
67
+ ### Kurationsgrund
68
+
69
+ <!-- Motivation for the creation of this dataset. -->
70
+ Der Datensatz wurde als Nebenprodukt eines Projekts zur Bereitstellung eines besseren Informationsabrufsystems für die öffentlichen Dienstleistungen der Landeshauptstadt München erstellt.
71
+ Der Datensatz wird regelmäßig aktualisiert, um neue Artikel aufzunehmen und die Einbettungen zu aktualisieren.
72
+
73
+ ### Quelldaten
74
+
75
+ <!-- This section describes the source data (e.g. news text and headlines, social media posts, translated sentences, ...). -->
76
+ Der Datensatz stammt von der Website der öffentlichen Dienstleistungen der Landeshauptstadt München und dem zugrunde liegenden CMS.
77
+ Alle Artikel wurden von Mitarbeitenden der Landeshauptstadt München geschrieben.
78
+
79
+ #### Datensammlung und -verarbeitung
80
+
81
+ <!-- This section describes the data collection and processing process such as data selection criteria, filtering and normalization methods, tools and libraries used, etc. -->
82
+ Die Daten wurden durch Abfragen einer internen API des CMS der Landeshauptstadt München gesammelt.
83
+ Die Daten wurden dann weiterverarbeitet, um die Metadaten und den Inhalt der Artikel zu extrahieren.
84
+ Der Artikelinhalt wurde mit der [markdownify](https://pypi.org/project/markdownify/) Bibliothek von HTML in Markdown umgewandelt.
85
+ Die Einbettungen wurden mit dem [`text-embedding-3-large`](https://platform.openai.com/docs/guides/embeddings/#embedding-models) Modell von OpenAI generiert.
86
+
87
+ #### Wer sind die Ersteller der Quelldaten?
88
+
89
+ <!-- This section describes the people or systems who originally created the data. It should also include self-reported demographic or identity information for the source data creators if this information is available. -->
90
+ Die Ersteller der Quelldaten sind die Mitarbeiter der Landeshauptstadt München, die die Artikel geschrieben haben, es gibt keine Zuordnung zu einzelnen Autoren.
91
+ Die Mitarbeiter sind eine vielfältige Gruppe von Menschen, die in der Stadtverwaltung arbeiten und über spezielles Wissen über die öffentlichen Dienstleistungen ihrer jeweiligen Referaten und Fähigkeiten im Verfassen verständlicher Artikel verfügen.
92
+
93
+ #### Persönliche und sensible Informationen
94
+
95
+ <!-- State whether the dataset contains data that might be considered personal, sensitive, or private (e.g., data that reveals addresses, uniquely identifiable names or aliases, racial or ethnic origins, sexual orientations, religious beliefs, political opinions, financial or health data, etc.). If efforts were made to anonymize the data, describe the anonymization process. -->
96
+ Der Datensatz enthält keine persönlichen oder sensiblen Informationen.
97
+
98
+ ## Verzerrungen, Risiken und Einschränkungen
99
+
100
+ <!-- This section is meant to convey both technical and sociotechnical limitations. -->
101
+ Der bereitgestellte Datensatz ist nur ein Teil aller öffentlichen Dienstleistungen der Landeshauptstadt München.
102
+ Darüber hinaus können die Daten Verzerrungen in Form von
103
+
104
+ - der in den Artikeln verwendeten Sprache (sogenanntes "Beamtendeutsch")
105
+ - der Qualität und Informationsfülle der Artikel
106
+ - der Anzahl der Artikel zu einem Thema
107
+ - und den mit den Artikeln verbundenen Schlüsselwörtern enthalten.
108
+
109
+ Dies kann zu Verzerrungen in den Einbettungen und dem auf dem Datensatz aufgebauten Systeme führen
110
+
111
+ ### Empfehlungen
112
+
113
+ <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
114
+ Um die Verzerrungen im Datensatz zu mindern, wird empfohlen, den Datensatz in Kombination mit anderen Informationsquellen zu verwenden.
115
+ Ein weiterer Ansatz besteht darin, den Datensatz mit einem benutzerdefinierten Evaluationsset zu bewerten, um Verzerrungen für den spezifischen Anwendungsfall zu identifizieren und zu mindern.
116
+
117
+ ## Glossar
118
+
119
+ - **Retrieval:** Abruf von Informationen aus einem Datensatz.
120
+ - **NLP:** Natural Language Processing, Verarbeitung natürlicher Sprache.
121
+ - **Embedding:** Einbettung, eine Darstellung von Text in Form von Vektoren.
122
+ - **LLM:** Large Language Model, großes Sprachmodell.
123
+ - **Desinformation:** Verbreitung falscher oder irreführender Informationen.
124
+
125
+ ## Weitere Informationen
126
+
127
+ Für weitere Informationen über die KI-Projekte der Landeshauptstadt München besuchen Sie bitte [ki.muenchen.de](https://ki.muenchen.de/).
128
+
129
+ ## Autoren der Dataset Card
130
+
131
+ Fabian Reinold - Landeshauptstadt München, it@M, KI Competence Center
132
+
133
+ ## Kontakt für die Dataset Card
134
+
135
+ Bei Fragen zum Datensatz kontaktieren Sie uns bitte [hier](mailto:[email protected]).