Dataset Viewer
Auto-converted to Parquet Duplicate
text
stringlengths
3
376
label
stringclasses
6 values
lang
stringclasses
15 values
id
stringlengths
7
13
Revenue Recognition
Tax & Accounting
English
Israel-4145
Más de la mitad de las empresas españolas fueron víctimas de algún delito económico en los últimos dos años
Government & Controls
Spanish
Spain-2044
Wynagrodzenie netto w Polsce to średnio 71% pensji brutto
Business & Management
Polish
Poland-1567
Time to talk: What has to change for women at work
Business & Management
English
Turkey-5447
Total Retail 2017
Industry
English
Spain-1981
European Private Business Survey 2018
Business & Management
English
Spain-2756
Növekvő vízügyi kockázatok
Industry
Hungarian
Hungary-61
Αναστολή συμβάσεων εργασίας Αυγούστου
Business & Management
Greek
Greece-10248
Zmiany w funkcjonowaniu rad nadzorczych zmierzają w dobrym kierunku lecz wymagają przyspieszenia
Business & Management
Polish
Poland-1339
Grundejer opnår tocifret millionbesparelse i grundskyld
Tax & Accounting
Danish
Denmark-12658
PwC refuerza su equipo directivo en el País Vasco con Aitor Soloeta y Estíbaliz Pérez
Business & Management
Spanish
Spain-2312
Compliance with Regulations and Standards
Government & Controls
English
Turkey-5410
Legal Flash: Social security of freelancers update
Finance
English
Greece-10775
Cumhurbaşkanı 21 sayılı, 15 Ağustos 2018 tarihli kararı Amerika Birleşik Devletleri Menşeli Bazı Ürünlerin İthalatında Ek Mali Yükümlülük Uygulanmasına Dair Kararda Değişiklik Yapılmasına İlişkin Karar
Tax & Accounting
Turkish
Turkey-7403
Jak się przygotować do wdrożenia nowego systemu IT?
Technology
Polish
Poland-668
Annual report 2016: Empowering our people transcript
Tax & Accounting
English
Turkey-9103
Ny tysk emballagelovgivning og risiko for store bøder
Government & Controls
Danish
Denmark-12678
NTT ComとPwCコンサルティング、デジタル設計データとAIを活用したデジタルマッチングプラットフォーム(設計/調達DX)を展開
Technology
Japanese
Japan-9869
Udsættelse af de nye fjernsalgsregler - hvad betyder det for dig?
Tax & Accounting
Danish
Denmark-13184
KDV ÖTV ve Tapu Harcı uygulamalarına ilişkin 287 sayılı Cumhurbaşkanı Kararı ÖTV ve Tapu Harcı uygulamalarına ilişkin 287 sayılı
Industry
Turkish
Turkey-7260
Avrupa borsalarında borç ihracı hakkında rehber
Finance
Turkish
Turkey-7025
Kundetilfredshed stryger til tops
Business & Management
Danish
Denmark-13538
Creating a strategic business plan and a development road map
Business & Management
English
Turkey-5314
Z Nesli Araştırması
Business & Management
Turkish
Turkey-6988
Working together to prevent diabetes in western Sydney
Industry
English
Spain-2821
La televisión y la publicidad en internet tirarán del sector de Entretenimiento y Medios en que crecerá un 29% hasta 2020
Industry
Spanish
Spain-2137
Podcast: Nyt om skat, uge 21 - deleøkonomisk aftale
Tax & Accounting
Danish
Denmark-12531
Κώδικας Φορολογικής Διαδικασίας (Νόμος 4174/2013)
Tax & Accounting
Greek
Greece-10065
Komşu teyzeler Eurobond günü yaparsa
Finance
Turkish
Turkey-8426
Implementering af IFRS 16
Tax & Accounting
Danish
Denmark-12549
Private equity and the responsible investment journey
Finance
English
Greece-10936
Bankacılık ve Sermaye Piyasaları
Finance
Turkish
Turkey-9541
Regnskabshåndbogen for pengeinstitutter
Tax & Accounting
Danish
Denmark-13562
Kim jest tester oprogramowania?
Technology
Polish
Poland-773
Digital - Transporte y Turismo
Industry
Spanish
Spain-1066
Bir iç soruşturma planı nasıl hazırlanmalı?
Government & Controls
Turkish
Turkey-6353
Lucas Palomar - Socio de Fusiones y adquisiciones
Finance
Spanish
Spain-746
Assistance with Capital Market Transactions
Finance
English
Israel-3955
Dahilde İşleme Rejimi Tebliği (İhracat: 2006/12)’nde Değişiklik Yapılmasına Dair Tebliğ
Tax & Accounting
Turkish
Turkey-8167
Communications Review: Is your organisation prepared for IFRS 9, 15, and 16?
Tax & Accounting
English
Turkey-5792
Deal summary AinaCom Oy
Finance
English
Finland-3058
Deal summary Vaaka Partners Oy / Unisport Scandinavia AB
Finance
English
Finland-3092
Insights from Sir Graeme Avery: Global Family Business Survey 2016: PwC
Business & Management
English
Greece-11097
KDV Genel Uygulama Tebliğinde Değişiklik Yapılmasına Dair 16 Seri No.lu Tebliğ
Tax & Accounting
Turkish
Turkey-8020
Regleringar den största risken inom energisektorn
Industry
Swedish
Sweden-14228
Yritysvastuubarometri 2016
Business & Management
Finnish
Finland-3449
Τραπεζική Αργία 2020
Finance
Greek
Greece-10355
Tax and legal news 15 December 2016
Tax & Accounting
English
Finland-3242
XIX Encuesta Mundial de CEOs - Retail y Consumo
Industry
Spanish
Spain-1446
Hisse senedi piyasaları
Finance
Turkish
Turkey-7023
Private Wealth
Finance
English
Finland-3322
Poradnik antykorupcyjny dla biznesu
Government & Controls
Polish
Poland-609
Quick fix: Købers momsnummer ved momsfrie EU-leverancer
Tax & Accounting
Danish
Denmark-12955
CEO Görüşleri
Business & Management
English
Turkey-6204
Industry 4.0,Industry 4.0,Industry 4.0,Industry 4.0,Industry 4.0
Technology
English
Spain-2773
Nedsættelse af sats for udbytteskat
Tax & Accounting
Danish
Denmark-13651
IFRS news - August 2016
Tax & Accounting
English
Greece-10731
CFOdirect: IFRS nyheder - nytårsfortsætter til din virksomhed, bliv forberedt på, hvilke spørgsmål du vil møde på aktionærmødet og meget mere
Tax & Accounting
Danish
Denmark-12225
Gümrük Genel Tebliği (Gümrük İşlemleri) (Seri No: 98)'nde Değişiklik (Seri No: 159)
Tax & Accounting
Turkish
Turkey-7590
C20 by Numbers 2017
Finance
English
Denmark-13375
Kira Geliri Elde Edenler İçin Hatırlatmalar
Industry
Turkish
Turkey-7173
Esimerkkejä yritysjärjestelytoimeksiannoistamme
Finance
Finnish
Finland-3708
Books
Industry
English
Italy-4921
XXII Encuesta Mundial de CEOs - PwC España
Business & Management
Spanish
Spain-1104
La fulgurante evolución de los eSports
Industry
Spanish
Spain-1617
Hábitos de consumo - 2018 Global Consumer Insights Survey
Industry
Spanish
Spain-1967
Young Women Network
Business & Management
English
Italy-4682
A régióban működő családi vállalkozások tulajdonosainak új generációja vezető szerepre készül
Business & Management
Hungarian
Hungary-171
ESMA Q&A vedr. guidelines til Alternative Performance Measures (APM)
Tax & Accounting
Danish
Denmark-12228
A1-attestationens praktiske værdi
Finance
Danish
Denmark-12177
Fremtidsutsikter for den norske mediebransjen
Industry
Norwegian
Norway-11820
Бизнес по-прежнему остается на стартовых позициях в вопросах интеграции и продвижения целей в области устойчивого развития
Business & Management
Russian
Russia-11693
Economic and business consulting
Business & Management
English
Israel-3929
Deal summary Lumicon Oy
Finance
English
Finland-3562
PwC Named a Leader in Global Cybersecurity Consulting and Global Digital Experience Agencies reports by Independent Research Firm
Technology
English
Global-13789
Key findings from the pharmaceuticals and life sciences industry
Industry
English
Turkey-5938
Luka VAT kosztuje kraje Europy Środkowo-Wschodniej 27 mld euro rocznie
Tax & Accounting
Polish
Poland-1428
Endnu en dansk regel er EU-stridig
Government & Controls
Danish
Denmark-12561
Private Business
Business & Management
English
Turkey-9199
Liikkeenjohdon konsultointi
Business & Management
Finnish
Finland-3484
İhraci kayda bağlı mallara ilişkin tebliğ (Tebliğ No: İhracat 2006/7)’de değişiklik yapılmasına dair tebliğ (İhracat: 2018/2)
Tax & Accounting
Turkish
Turkey-7443
Kira gelirlerinin beyanı
Industry
Turkish
Turkey-8685
İthalatta Haksız Rekabetin Önlenmesine İlişkin Tebliğ (Tebliğ No: 2018/15)
Tax & Accounting
Turkish
Turkey-7478
Fondshåndbogen 2017
Finance
Danish
Denmark-13445
La gestione dell’informativa interna
Tax & Accounting
Italian
Italy-4798
Improve operational efficiency
Business & Management
English
Finland-2976
Balázs Gábor: IFRS - Komoly számviteli változások jöhetnek
Tax & Accounting
Hungarian
Hungary-325
Yağlı Ayçiçeği Tohumu İthalatinda Tarife Kontenjanı Uygulanması Hakkında Karar (Karar Sayı 497)
Tax & Accounting
Turkish
Turkey-7562
Deal summary Oy Kaha Ab
Finance
English
Finland-3549
Benzin ve Motorinde ÖTV İndirimi
Tax & Accounting
Turkish
Turkey-7897
Monitoring, Reporting and Validating (MRV) maritime vessels’ CO2 emissions performance: A new era in the Shipping Industry
Industry
English
Greece-10699
一般社団法人日本トップリーグ連携機構とPwCコンサルティング、女性スポーツリーグ同士が連携するWoman Athletes Projectを発足
Industry
Japanese
Japan-9779
Real Estate Compliance
Industry
English
Greece-11251
PwC’nin son raporuna göre: Kadınlar iş yaşamında yükselmeye hazır ve işverenlerden şeffaflık bekliyorlar
Business & Management
Turkish
Turkey-6028
PwCコンサルティングとテラドローン、空飛ぶクルマ/ドローンの現場活用支援で協業を開始
Technology
Japanese
Japan-9708
2019 Bankacılık ve sermaye piyasaları
Finance
Turkish
Turkey-9594
Baselkomiteen opretter ny klimaenhed
Business & Management
Danish
Denmark-13056
Chairman's message: Global Annual Review 2020
Business & Management
English
Spain-1235
Firmy będą musiały skutecznie zapobiegać nadużyciom
Government & Controls
Polish
Poland-759
עדכון שיעור הריבית לעניין סעיפים 3(ט) ו-3(י) לפקודת מס הכנסה לשנת 2013
Tax & Accounting
Hebrew
Israel-4384
End of preview. Expand in Data Studio

awinml/MultiFin(複製)

データの内容

15 言語の実際の記事の見出しに、階層的なトピックのラベルを付けた多言語のデータ。1 行が 1 件の見出しで、text に見出し、lang に言語、label にトピックが入る。ラベルは母語話者による付与と、訳してから付与する方法で作られた。日本語は 230 件と少なく、利用条件は非商用(NC)である。

データの概要

項目 値
行数 37,848(6 構成・14 分割の合計)
列数 4
容量 2.14 MB(ファイル 16 本)
ファイルの形式 parquet 14
言語 en, es, pl, hu, el, da, tr, ja, sv, fi, no, ru, it, he, is
公開 public

構成と分割

構成(config) 分割(split)ごとの行数 列数
all_languages_highlevel train 6,430、validation 1,608、test 2,010 4
all_languages_lowlevel train 6,430、validation 1,608、test 2,010 4
high_resources_highlevel train 5,353、validation 1,339 4
high_resources_lowlevel train 5,353、validation 1,339 4
only_english_highlevel train 1,747、validation 437 4
only_english_lowlevel train 1,747、validation 437 4

列(構成 all_languages_highlevel・分割 train。構成ごとに列が違う)

列 型 先頭の行の値
text string Revenue Recognition
label string Tax & Accounting
lang string English
id string Israel-4145

行数と列は HF のデータセットビューアの集計(元のデータセットの現在の版で数えたもの)。全データセットの一覧は HYUKI5/dataset-catalog にある。

複製について

awinml/MultiFin をリビジョン d44a842791acb83a104440a5662b787537ff980b の時点でそのまま複製したもの。中身は変えていない。判断モデル Laya(convaiinnovations/laya)の微調整に使う候補として、調査メモ(Laya 学習用データセット調査)に載せたデータを集めている。

出典

論文

アノテーションの方法

(元のカードに該当の見出しが無いので、ラベルの付け方に触れた段落を抜き出した)

MULTIFIN データセットは、15 言語の実際の記事の見出しからなる多言語コーパスである。コーパスは階層的なラベル構造を用いてアノテーションされ、マルチクラス分類とマルチラベル分類の二つの分類タスクを提供する。

本データセットは 15 言語の 10,048 件の見出しからなり、下位レベル(LOW-LEVEL)の 23 個のトピックラベルと、マルチクラス用の上位レベル(HIGH-LEVEL)の 6 個のトピックでアノテーションされている。

微調整に使うときの注意

  • 調査メモの台帳での区分: C(NC・ND・研究専用・評価専用など、商用の利用を妨げうる記載がある)。区分は記載の文言で機械的に付けたもので、法的な判定ではない。学習に入れる前に元のカードとライセンスの原文で確かめる。
  • Laya の入力は 2,048 token まで(ModernBERT-large / mmBERT)。長い本文は先頭から黙って切られるので、切れる行は事前に除くか、要約・切り出しをしてから入れる。
  • 分割は記事・文書単位で行い、同じ本文から作った別の問いを学習側と較正側に分けない。温度と閾値は学習に使っていない分割で別々に当てる。

調査メモの記述

メモ 節 ライセンス(記載) 確認 注記
japanese_finance_disclosure.md Q3. 日本語の金融文書で、感情・トピック・「約束と回避」のラベルが付いた小〜中規模のデータは何か(chABSA・MultiFin-ja・JF-ICR など) (L80) CC BY-NC 4.0(JFinTEB README)。awinml/MultiFin の license 欄は空 確認済 原文注記: 日本語分は 230 件と小さい。NC。MultiFin の英語版は既出(finance_news.md)

元のカード(日本語訳)

以下は元のデータセットのカードの本文を日本語に訳したもの(機械翻訳)。訳に疑問があるときは、末尾の原文を確かめる。


MultiFin

MultiFin は、異なる文字体系と語族にまたがる 15 言語の、実際の記事の見出しからなる公開されている金融データセットである。 本データセットは階層的なラベル構造を持ち、マルチラベルとマルチクラスの二つの分類タスクを提供する。

データセットの説明

MULTIFIN データセットは、15 言語の実際の記事の見出しからなる多言語コーパスである。コーパスは階層的なラベル構造を用いてアノテーションされ、マルチクラス分類とマルチラベル分類の二つの分類タスクを提供する。

  • 作成者: Rasmus Jørgensen, Oliver Brandt, Mareike Hartmann, Xiang Dai, Christian Igel, and Desmond Elliott.
  • 言語 (NLP): 英語、スペイン語、ポーランド語、ハンガリー語、ギリシャ語、デンマーク語、トルコ語、日本語、スウェーデン語、フィンランド語、ノルウェー語、ロシア語、イタリア語、ヘブライ語、アイスランド語。
  • ライセンス: [More Information Needed]

データセットの出典

データセットの構造

本データセットは 15 言語の 10,048 件の見出しからなり、下位レベル(LOW-LEVEL)の 23 個のトピックラベルと、マルチクラス用の上位レベル(HIGH-LEVEL)の 6 個のトピックでアノテーションされている。

さらに、データセットは次の二つの部分集合に層化されている:

  1. only_english: 英語の訓練データのみを含む。
  2. high_resources: 資源の豊富な 5 言語(英語、トルコ語、デンマーク語、スペイン語、ポーランド語)を含む部分集合。

引用

BibTeX:

@inproceedings{jorgensen-etal-2023-multifin,
    title = "{M}ulti{F}in: A Dataset for Multilingual Financial {NLP}",
    author = "J{\o}rgensen, Rasmus  and
      Brandt, Oliver  and
      Hartmann, Mareike  and
      Dai, Xiang  and
      Igel, Christian  and
      Elliott, Desmond",
    editor = "Vlachos, Andreas  and
      Augenstein, Isabelle",
    booktitle = "Findings of the Association for Computational Linguistics: EACL 2023",
    month = may,
    year = "2023",
    address = "Dubrovnik, Croatia",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2023.findings-eacl.66",
    doi = "10.18653/v1/2023.findings-eacl.66",
    pages = "894--909",
    abstract = "Financial information is generated and distributed across the world, resulting in a vast amount of domain-specific multilingual data. Multilingual models adapted to the financial domain would ease deployment when an organization needs to work with multiple languages on a regular basis. For the development and evaluation of such models, there is a need for multilingual financial language processing datasets. We describe MultiFin {--} a publicly available financial dataset consisting of real-world article headlines covering 15 languages across different writing systems and language families. The dataset consists of hierarchical label structure providing two classification tasks: multi-label and multi-class. We develop our annotation schema based on a real-world application and annotate our dataset using both {`}label by native-speaker{'} and {`}translate-then-label{'} approaches. The evaluation of several popular multilingual models, e.g., mBERT, XLM-R, and mT5, show that although decent accuracy can be achieved in high-resource languages, there is substantial room for improvement in low-resource languages.",
}
原文(英語)

MultiFin

MultiFin – a publicly available financial dataset consisting of real-world article headlines covering 15 languages across different writing systems and language families. The dataset consists of hierarchical label structure providing two classification tasks: multi-label and multi-class.

Dataset Description

The MULTIFIN dataset is a multilingual corpus, consisting of real-world article headlines covering 15 languages. The corpus is annotated using hierarchical label structure, providing two classification tasks: multi-class and multi-label classification.

  • Curated by: Rasmus Jørgensen, Oliver Brandt, Mareike Hartmann, Xiang Dai, Christian Igel, and Desmond Elliott.
  • Language(s) (NLP): English, Spanish, Polish, Hungarian, Greek, Danish, Turkish, Japanese, Swedish, Finnish, Norwegian, Russian, Italian, Hebrew, Icelandic.
  • License: [More Information Needed]

Dataset Sources

Dataset Structure

The dataset consists of 10,048 headlines in 15 languages annotated with 23 topic labels for LOW-LEVEL and 6 HIGH-LEVEL topics for multi-class.

The dataset has been further stratified into two subsets:

  1. only_english: that contains only English training data.
  2. high_resources: a subset that contains 5 high-resource languages (i.e., English, Turkish, Danish, Spanish, Poland).

Citation

BibTeX:

@inproceedings{jorgensen-etal-2023-multifin,
    title = "{M}ulti{F}in: A Dataset for Multilingual Financial {NLP}",
    author = "J{\o}rgensen, Rasmus  and
      Brandt, Oliver  and
      Hartmann, Mareike  and
      Dai, Xiang  and
      Igel, Christian  and
      Elliott, Desmond",
    editor = "Vlachos, Andreas  and
      Augenstein, Isabelle",
    booktitle = "Findings of the Association for Computational Linguistics: EACL 2023",
    month = may,
    year = "2023",
    address = "Dubrovnik, Croatia",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2023.findings-eacl.66",
    doi = "10.18653/v1/2023.findings-eacl.66",
    pages = "894--909",
    abstract = "Financial information is generated and distributed across the world, resulting in a vast amount of domain-specific multilingual data. Multilingual models adapted to the financial domain would ease deployment when an organization needs to work with multiple languages on a regular basis. For the development and evaluation of such models, there is a need for multilingual financial language processing datasets. We describe MultiFin {--} a publicly available financial dataset consisting of real-world article headlines covering 15 languages across different writing systems and language families. The dataset consists of hierarchical label structure providing two classification tasks: multi-label and multi-class. We develop our annotation schema based on a real-world application and annotate our dataset using both {`}label by native-speaker{'} and {`}translate-then-label{'} approaches. The evaluation of several popular multilingual models, e.g., mBERT, XLM-R, and mT5, show that although decent accuracy can be achieved in high-resource languages, there is substantial room for improvement in low-resource languages.",
}
Downloads last month
82