CoFiF Plus: A French Financial Narrative Summarisation Corpus
16043Authors
nan
Abstract
AbstractNatural Language Processing is increasingly being applied in the finance and business industry to analyse the text of many different types of financial documents. Given the increasing growth of firms around the world, the volume of financial disclosures and financial texts in different languages and forms is increasing sharply and therefore the study of language technology methods that automatically summarise content has grown rapidly into a major research area. Corpora for financial narrative summarisation exists in English, but there is a significant lack of financial text resources in the French language. To remedy this, we present CoFiF Plus, the first French financial narrative summarisation dataset providing a comprehensive set of financial text written in French. The dataset has been extracted from French financial reports published in PDF file format. It is composed of 1,703 reports from the most capitalised companies in France (Euronext Paris) covering a time frame from 1995 to 2021. This paper describes the collection, annotation and validation of the financial reports and their summaries. It also describes the dataset and gives the results of some baseline summarisers. Our datasets will be openly available upon the acceptance of the paper.
Keywords
Full Content
Proceedings of the 13th Conference on Language Resources and Evaluation (LREC 2022), pages 1622–1639 Marseille, 20-25 June 2022 © European Language Resources Association (ELRA), licensed under CC-BY-NC-4.0 1622 CoFiF Plus: A French Financial Narrative Summarisation Corpus Nadhem Zmandar†, Tobias Daudert, Sina Ahmadi∗, Mahmoud El-Haj†, Paul Rayson† †UCREL NLP group, School of Computing and Communications, Lancaster University, UK ∗Insight SFI Research Centre for Data Analytics, National University of Ireland Galway, Ireland {n.zmandar, m.el-haj, p.rayson}@lancaster.ac.uk t.daudert@gmail.com, sina.ahmadi@insight-centre.org Abstract Natural Language Processing is increasingly being applied in the finance and business industry to analyse the text of many different types of financial documents. Given the increasing growth of firms around the world, the volume of financial disclosures and financial texts in different languages and forms is increasing sharply and therefore the study of language technology methods that automatically summarise content has grown rapidly into a major research area. Corpora for financial narrative summarisation exist in English, but there is a significant lack of financial text resources in the French language. To remedy this, we present CoFiF Plus, the first financial narrative summarisation dataset providing a comprehensive set of financial text written in the French language. The dataset has been extracted from french financial reports published in PDF file format. It is composed of 1,703 reports from the most capitalised companies in France (Euronext Paris) covering a time frame from 1995 to 2021. This paper describes the collection, annotation and validation of the financial reports and their summaries. It also describes the dataset and gives the results of some baseline summarisers. Keywords: Summarisation, French financial narratives, French NLP, corpus, dataset construction 1. Introduction Information exchange is crucial in any market and forms the foundation for participants to act on. In a financial setting, this ensures transparency and con- tributes to investors’ confidence while depicting the credibility and quality of a financial marketplace (PwC France, 2021). Listed companies are legally obligated to communicate regularly with their shareholders. The financial communication policy reflects the regulatory requirements related to going public as well as the willingness of the executives to regularly communicate with financial markets in a transparent, professional and responsive fashion. Hence, firms use narratives to communicate with investors, shareholders, employees, analysts, regulators and rating agencies among others. Companies aim to create a relationship of trust with the market by using these communications as a reliable source of pertinent information. Stakeholders can then assess a company’s ability to create value and judge its prospects. With a growing plethora of digitised finan- cial communications, novel approaches for synthesis- ing relevant information as well as aggregating those into digestible forms are needed. No investor can read thousands of documents yearly containing up to hun- dreds of pages. As such, summaries play a key role in communicating written information. Defined as a document reduced only to its essential content, a sum- mary helps readers to understand the material more ef- fectively. Summaries also play a pivotal role in direct- ing readers’ attention to relevant parts of a document. The …
Content truncated to 500 words
Labeled Text
AbstractNatural Language Processing is increasingly being applied in the finance and business industry to analyse the text of many different types of financial documents. Given the increasing growth of firms around the world, the volume of financial disclosures and financial texts in different languages and forms is increasing sharply and therefore the study of language technology methods that automatically summarise content has grown rapidly into a major research area. Corpora for financial narrative summarisation exists in English, but there is a significant lack of financial text resources in the French language. To remedy this, we present CoFiF Plus, the first French financial narrative summarisation dataset providing a comprehensive set of financial text written in French. The dataset has been extracted from French financial reports published in PDF file format. It is composed of 1,703 reports from the most capitalised companies in France (Euronext Paris) covering a time frame from 1995 to 2021. This paper describes the collection, annotation and validation of the financial reports and their summaries. It also describes the dataset and gives the results of some baseline summarisers. Our datasets will be openly available upon the acceptance of the paper.
Publication Details
- Paper ID:
- 16043
- Year:
- 2022
- Publication:
- TIPSTER
- Publication Type:
- conf
- DOI:
- nan