Hugging Face's logo

Datasets:
slm-ct
/
fas_ad_practice_dataset

Tasks:
Text Classification
Feature Extraction
Modalities:
Text
Formats:
json
Languages:
Russian
Size:
< 1K
Tags:
legal
advertising
antimonopoly
caselaw
russian
Libraries:
Datasets
pandas
License:
Dataset card Data Studio Files
xet
Community
1
fas_ad_practice_dataset / data_dictionary.md
slm-ct's picture
slm-ct
Upload 6 files
8d64453 verified
preview code
|
raw
history blame contribute delete
6.06 kB

Data Dictionary

Dataset Fields Description

Note: The examples provided are in Russian as they are taken directly from the source data. An English translation is provided in parentheses for clarity.

1. docId

  • Type: String (UUID)
  • Description: A unique document identifier in UUID format.
  • Example: 0009fdfc-f2c8-4ab7-9562-9e5da0e9f102
  • Required: Yes

2. Violation_Type

  • Type: String (Categorical)
  • Description: The type of advertising law violation.
  • Possible Values:
    • substance - violations related to the content of the advertisement (e.g., misleading, unethical).
    • placement - violations related to the rules of ad distribution (e.g., location, time, method).
  • Required: Yes

3. document_date

  • Type: Date (YYYY-MM-DD)
  • Description: The date the FAS decision was made.
  • Format: ISO 8601 (YYYY-MM-DD)
  • Example: 2021-09-07
  • Required: Yes

4. FASbd_link

  • Type: URL
  • Description: A direct link to the case decision in the official FAS Russia database.
  • Example: https://br.fas.gov.ru/cases/4d3ec9bc-68ec-40c8-8187-4c5d47909382/
  • Required: Yes

5. FAS_division

  • Type: String
  • Description: The regional division of the FAS that handled the case.
  • Examples:
    • Московское УФАС (Moscow UFAS)
    • Санкт-Петербургское УФАС (Saint Petersburg UFAS)
    • ФАС России (FAS Russia (central office))
  • Required: Yes

6. violation_found

  • Type: String
  • Description: Indicates whether a violation was found.
  • Possible Values:
    • да - violation found (yes)
    • нет - violation not found (no)
  • Required: Yes

7. defendant_name

  • Type: String
  • Description: The name of the entity held liable (advertiser or ad distributor).
  • Examples: ООО "КТЛ" (LLC "KTL"), ИП Иванов И.И. (IE Ivanov I.I.)
  • Required: Yes

8. defendant_industry

  • Type: String
  • Description: The industry/sector of the entity held liable.
  • Examples:
    • Розничная торговля (Retail Trade)
    • Финансовые услуги (Financial Services)
    • Медицинские услуги (Medical Services)
  • Required: Yes

9. ad_description

  • Type: Text
  • Description: A detailed description of the advertisement, including its format, visual elements, and context.
  • Example: Реклама услуг комиссионного магазина... (Advertisement for a consignment store's services...)
  • Required: Yes

10. ad_content_cited

  • Type: Text
  • Description: A direct quote from the advertisement's text that was the subject of the review.
  • Example: "Срочный выкуп техники по максимальной цене в городе!" ("Urgent buyout of electronics at the best price in town!")
  • Required: No (may be absent for visual-only ads).

11. ad_platform

  • Type: String
  • Description: The platform or channel where the ad was distributed.
  • Examples:
    • Наружная реклама (Outdoor advertising)
    • Интернет-сайт (Internet website)
    • Телевидение (Television)
  • Required: Yes

12. violation_summary

  • Type: Text
  • Description: A brief summary of the nature of the violation.
  • Example: Использование превосходной степени без указания конкретного критерия сравнения (Use of a superlative claim without specifying a concrete comparison criterion)
  • Required: Yes

13. FAS_arguments

  • Type: Text
  • Description: The detailed argumentation from the FAS, including:
    • Analysis of the advertising materials.
    • References to expert opinions.
    • Interpretation of legal norms.
    • Assessment of the evidence presented.
  • Required: Yes

14. legal_provisions

  • Type: String (List)
  • Description: The violated norms of the Federal Law "On Advertising".
  • Format: Comma-separated articles and clauses.
  • Examples:
    • ч. 1 ст. 5 - недостоверная реклама (Part 1 Art. 5 - misleading advertising)
    • ч. 2 ст. 5 - недобросовестная реклама (Part 2 Art. 5 - unfair advertising)
    • ст. 21 - реклама алкоголя (Art. 21 - alcohol advertising)
    • ст. 24 - реклама лекарств (Art. 24 - advertising of medicines)
  • Required: Yes

15. thematic_tags

  • Type: String (List)
  • Description: Thematic tags for categorizing cases.
  • Example Tags:
    • некорректные сравнения (improper_comparisons)
    • БАД_дисклеймер (dietary_supplement_disclaimer)
    • запрет_умолчаний (omission_of_material_information)
  • Format: Comma-separated tags.
  • Required: No

Data Characteristics

Handling of Missing Values

  • Null values may occur in the ad_content_cited and thematic_tags fields.
  • All other fields are mandatory.

Encoding

  • UTF-8 for all text fields.
  • CSV Delimiter: ; (semicolon).

Time Period

  • The data covers the period from 2019 to 2025.

Regional Distribution

  • All federal districts of Russia are represented.
  • Approximately 30% of cases are decisions from the central office of the FAS.
  • 70% are decisions from regional divisions.

Usage Notes

  1. For Legal Analysis: The legal_provisions field allows for quick filtering of cases by specific articles of the law.

  2. For ML Models: The combination of ad_description + ad_content_cited provides a comprehensive representation of the advertising material.

  3. For Statistics: The defendant_industry and ad_platform fields enable the analysis of industry-specific trends.

  4. For Temporal Analysis: The document_date field in ISO format facilitates easy time-series analysis to identify trends in regulatory activity.