The 1st Joint Workshop on Financial Narrative Processing and MultiLing Financial Summarisation (FNP-FNS 2020)

FinCausal-2020 Shared Task: “Financial Document Causality Detection”

To be held at The 28th International Conference on Computational Linguistics (COLING’2020), Barcelona, Spain [online] on 12 December 2020.

FNP-FNS Online Running Instructions:

Workshop Program: Click here to see the workshop schedule

Keynote speaker: Dr Ana Gisbert, to join the talk:


TASK 1 – Leader board 

Team F1 Score Recall Precision
LIORI 97.75 (1) 97.77 (1) 97.73 (1)
UPB 97.55 (2) 97.59 (2) 97.53 (2)
ProsperAMnet 97.23 (3) 97.20 (3) 97.28 (3)
FiNLP 96.99 (4) 97.03 (4) 96.96 (4)
DOMINO 96.12 (5) 96.06 (5) 96.19 (5)
IITkgp 95.78 (6) 95.83 (6) 95.74 (6)
LangResearchLab_NC 95.00 (7) 94.92 (7) 95.08 (7)
NITK NLP 94.35 (8) 94.87 (8) 94.32 (8)
Fraunhofer IAIS 94.29 (9) 94.76 (9) 94.20 (9)
ISIKUN 93.09 (10) 94.33 (10) 93.89 (10)

TASK 2 – Leader board 

Team F1 Score Recall Precision Exact match
NTUNLPL 94.72 (1) 94.70 (1) 94.79 (1) 82.45 (1)
GBe 94.66 (2) 94.66 (2) 94.67 (2) 73.67 (2)
ProsperAMnet 83.71 (3) 83.63 (3) 83.92 (3) 70.38 (4)
LIORI 82.60 (4) 82.80 (4) 82.48 (4) 70.53 (3)
DOMINO 79.60 (5) 78.90 (5) 81.90 (5) 00.00 (7)
Fraunhofer IAIS 76.00 (6) 74.89 (7) 79.95 (6) 19.12 (5)
JDD 75.61 (7) 75.57 (6) 75.95 (7) 00.00 (7)
UPB 73.10 (8) 72.14 (8) 75.61 (8) 18.34 (6)

Provisional Key Dates:

  • Trial data set released on the 1st of February 2020
  • Training data released on the 1st of March 2020
  • Blind test dataset released on the 1st of May 2020.
  • Contributions from participants are expected on the 1st of June 2020 -> 24th of June 2020
  • Release of results are provided by organizers on June 8, 2020 -> 1st of July 2020
  • Shared task papers due September 1, 2020
  • Notification of acceptance October 1, 2020
  • Camera-ready papers due November 1, 2020
  • Workshop and shared task dates December 12, 2020

Submission guidelines:

Participation Form: Register a new team


Financial analysis needs factual data, but also explanation on the variability of these data. Data state facts, but provide little to no knowledge regarding how these facts materialised. The Financial Document Causality Detection Task aims to develop an ability to explain, from external sources, the reasons why a transformation occurs in the financial landscape, as a preamble to generating accurate and meaningful financial narrative summaries. Its goal is to evaluate which events or which chain of events can cause a financial object to be modified or an event to occur, regarding a given external context. This context is available in the financial news, but due to the high volatility of such information, mapping an external cause to a given consequence is not trivial.

The task dataset has been extracted from different 2019 financial news kindly provided by Qwam, and additional SEC data from the Edgar Database, and has been normalised for the research task.

Participants will be asked to evaluate whether a sentence is causal or not (Task 1), then to detect, in causal sentences, which elements of the sentence relate to the cause and which relate to the effect (Task 2).

This paper details the data processing and the labelling scheme, the expected results and the metrics used for evaluation. It will be updated on release of the training data.


As part of the Financial Narrative workshop, we propose the FinCausal Task, focusing on detecting if an object, an event or a chain of events is considered a cause for a prior event.  This shared task focuses on determining causality associated to a quantified fact. An event is defined as the arising or emergence of a new object or context in regard of a previous situation. So the task will emphasise the detection of causality associated to transformation of financial objects embedded in quantified facts.

Participants will be provided with a sample of text blocks extracted from financial news and SEC data, labelled through inter annotator agreement.

 The Shared Task contains two sub-tasks:

Task 1: Sentence Classification

This task is a binary classification task. The goal of this subtask is to filter sentences which display causal meanings (1) from the sentences that are noise in regard of causality (0)

Table 1: Sentence Classification Sample



As customer expectations continuously evolve, customers expect immediacy and simplicity.


Thomas Cook’s subsidiary in Germany is still technically operating as of Monday afternoon but has stopped taking bookings. More than 140,000 German holidaymakers have been impacted and tens of thousands of future travel bookings may not be honored


According to Gran , the company has no plans to move all production to Russia , although that is where the company is growing



Task 2: Cause and Effect Detection

This task is a relation detection task. The aim is to identify, in a causal sentence or text block, the causal elements and the consequential ones.

Table 2: Cause and Effect Detection Sample




Boussard Gavaudan Investment Management LLP bought a new position in shares of GENFIT S A/ADR in the second quarter worth about $199,000. Morgan Stanley increased its stake in shares of GENFIT S A/ADR by 24.4% in the second quarter.Morgan Stanley now owns 10,700 shares of the company’s stock worth $211,000 after purchasing an additional 2,100 shares during the period

Morgan Stanley increased its stake in shares of GENFIT S A/ADR by 24.4% in the second quarter

Morgan Stanley now owns 10,700 shares of the company’s stock worth $211,000 after purchasing an additional 2,100 shares during the period.

Zhao found himself 60 million yuan indebted after losing 9,000 BTC in a single day (February 10, 2014)

losing 9,000 BTC in a single day (February 10, 2014)

Zhao found himself 60 million yuan indebted


Participants are free to use any method they see fit (regex, corpus linguistics, entity relationship models, deep learning methods) to identify the causes and effects.

Shared task Paper Submission Instructions:

Submission URL:

Detailed submission guidelines can be found here:


Shared Task Organisers

  • Dominique Mariko – Yseop Lab
  • Hanna Abi Akl – Yseop Lab
  • Hugues de Mazancourt – Yseop Lab
  • Estelle Labidurie – Yseop Lab
  • Stephane Durfort – Yseop Lab


Shared Task Contact 

The participants to this task will access the data after registering, and thereby pledge to contribute to the workshop by submitting an experiment paper.

Participant can register to this shared task by filling this form, and get access to the datasets.

For any question please contact the organisers at