Event Causality Extraction from Natural Science Literature

Barik, Biswanath; Marsi, Erwin; Öztürk, Pınar

doi:10.13053/rcs-117-1-8

Cited by 9 publications

(6 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Meanwhile, unlike non-statistical architectures that only extracted explicit cause-effect relations, a large number of ML systems (e.g., [9,80,83,92]) have the capability to explore implicit relations. In the same year with the study of [2], Barik et al [4] categorize existing CE approaches into four groups: using handcrafted patterns, using semiautomatic causal patterns, using supervised learning, and statistical methods. From their point of view, instead of using manually linguistic clues and domain knowledge, semiautomatic learning acquires lexico-syntactic patterns from a larger corpus automatically.…”

Section: Previous Surveysmentioning

confidence: 99%

A survey on extraction of causal relations from natural language text

Han

Poon

2022

Knowl Inf Syst

View full text Add to dashboard Cite

As an essential component of human cognition, cause–effect relations appear frequently in text, and curating cause–effect relations from text helps in building causal networks for predictive tasks. Existing causality extraction techniques include knowledge-based, statistical machine learning (ML)-based, and deep learning-based approaches. Each method has its advantages and weaknesses. For example, knowledge-based methods are understandable but require extensive manual domain knowledge and have poor cross-domain applicability. Statistical machine learning methods are more automated because of natural language processing (NLP) toolkits. However, feature engineering is labor-intensive, and toolkits may lead to error propagation. In the past few years, deep learning techniques attract substantial attention from NLP researchers because of its powerful representation learning ability and the rapid increase in computational resources. Their limitations include high computational costs and a lack of adequate annotated training data. In this paper, we conduct a comprehensive survey of causality extraction. We initially introduce primary forms existing in the causality extraction: explicit intra-sentential causality, implicit causality, and inter-sentential causality. Next, we list benchmark datasets and modeling assessment methods for causal relation extraction. Then, we present a structured overview of the three techniques with their representative systems. Lastly, we highlight existing open challenges with their potential directions.

show abstract

Section: Previous Surveysmentioning

confidence: 99%

A survey on extraction of causal relations from natural language text

Han

Poon

2022

Knowl Inf Syst

View full text Add to dashboard Cite

show abstract

“…A causal relation encodes a semantic relationship between two arguments, where one is the Cause argument, and the other is the Effect argument, in which the occurrence of the Cause leads to the occurrence of the Effect (Barik et al, 2016). A Cause can be a reason, explanation or justification that leads to an Effect (Webber et al, 2019).…”

Section: Introductionmentioning

confidence: 99%

RECESS: Resource for Extracting Cause, Effect, and Signal Spans

Tan,

Hettiarachchi,

Hürriyetoğlu

et al. 2023

Proceedings of the 13th International Joint Conference on Natural Language Processing and the 3rd Conference of the Asia-Pacifi

View full text Add to dashboard Cite

Causality expresses the relation between two arguments, one of which represents the cause and the other the effect (or consequence). Causal relations are fundamental to human decision making and reasoning, and extracting them from natural language texts is crucial for building effective natural language understanding models. However, the scarcity of annotated corpora for causal relations poses a challenge in the development of such tools. Thus, we created Resource for Extracting Cause, Effect, and Signal Spans (RECESS), a comprehensive corpus annotated for causality at different levels, including Cause, Effect, and Signal spans. The corpus contains 3,767 sentences, of which, 1,982 are causal sentences that contain a total of 2,754 causal relations. We report baseline experiments on two natural language tasks (Causal Sentence Classification, and Cause-Effect-Signal Span Detection), and establish initial benchmarks for future work. We conduct an in-depth analysis of the corpus and the properties of causal relations in text. RECESS is a valuable resource for developing and evaluating causal relation extraction models, benefiting researchers working on topics from information retrieval to natural language understanding and inference.

show abstract

“…Causal relation extraction methods identify the connections within the text and then aggregate them (Asghar, 2016;Bach & Badaskar, 2007;Khoo & Na, 2006). The latter is generally more appropriate for evidence synthesis as its bottom-up nature covers even infrequently occurring variables and retains the stated relationships among variables (Barik et al, 2016). Causal relation extraction techniques can, in turn, be classi ed into knowledge-based, statistical machine learning, and deep learning techniques (Yang et al, 2021).…”

Section: Introductionmentioning

confidence: 99%

A semi-automated approach to policy-relevant evidence synthesis: Combining natural language processing, causal mapping, and graph analytics for public policy

Hooper,

Goyal,

Blok

et al. 2023

Preprint

View full text Add to dashboard Cite

Although causal evidence synthesis is critical for the policy sciences – whether it be analysis for policy or analysis of policy – its repeatable, systematic, and transparent execution remains challenging due to the growing volume, variety, and velocity of policy-relevant evidence generation as well as the complex web of relationships within which policies are usually situated. To address these shortcomings, we developed a novel, semi-automated approach to synthesizing causal evidence from policy-relevant documents. Specifically, we propose the use of natural language processing (NLP) for the extraction of causal evidence and subsequent homogenization or normalization of the varied text, causal mapping for the collation, visualization, and summarization of complex interdependencies within the policy system, and graph analytics for further investigation of the structure and dynamics of the causal map. We illustrate this approach by applying it to a collection of 28 articles on the emissions trading scheme (ETS), a policy instrument of increasing importance for climate change mitigation. In all, we find 300 variables and 284 cause-effect pairs in our input dataset (consisting of 4524 sentences), which are reduced to 70 unique variables and 119 cause-effect pairs after normalization. We create a causal map depicting these and analyze it subsequently to obtain systemic perspective as well as policy-relevant insight on the ETS that is broadly consistent with select manually conducted, previous meta-reviews of the policy instrument. We conclude that, despite its present limitations, this approach can help synthesize causal evidence for policy analysis, policymaking, and policy research.

show abstract

Event Causality Extraction from Natural Science Literature

Cited by 9 publications

References 27 publications

A survey on extraction of causal relations from natural language text

A survey on extraction of causal relations from natural language text

RECESS: Resource for Extracting Cause, Effect, and Signal Spans

A semi-automated approach to policy-relevant evidence synthesis: Combining natural language processing, causal mapping, and graph analytics for public policy

Contact Info

Product

Resources

About