Papers by Mark Finlayson
Distinguishing Between Foreground and Background Events in News (2020.coling-main)
Copied to clipboard
| Challenge: | a new task is needed to distinguish between foreground and background events in news articles . |
| Approach: | They propose a task of distinguishing between foreground and background events in news articles . they also identify the general temporal position of background events relative to the foregoing period . |
| Outcome: | The proposed model achieves good performance on a dataset of news articles . |
Detecting Subevents using Discourse and Narrative Features (P19-1)
Copied to clipboard
| Challenge: | Existing models for detecting events as subevents have been developed for analyzing textual understanding. |
| Approach: | They propose a supervised model that automatically identifies when one event is a subevent of another. |
| Outcome: | The proposed model outperforms previous systems on two annotated corpora with event hierarchies, achieving 0.74 BLANC F1 on the Intelligence Community corpus and 0.70 F1 for the HiEve corpus, respectively a 15 and 5 percentage point improvement over previous models. |
A Comprehensive Evaluation and Correction of the TimeBank Corpus (2022.lrec-1)
Copied to clipboard
| Challenge: | TimeML is an annotation scheme for capturing temporal information in text. |
| Approach: | They propose to use TimeML to validate TimeML and provide a rich dataset of events, temporal expressions, and temporal relationships for training and testing temporal analysis systems. |
| Outcome: | The proposed methods detect and correct errors in the TimeML corpus and provide a reference corpus for training and testing temporal analysis systems. |
jTLEX: a Java Library for TimeLine EXtraction (2023.eacl-demo)
Copied to clipboard
| Challenge: | Timeline EXtraction library provides Java implementation of TimeML annotations and tools for programmatic manipulation of Timeline graphs. |
| Approach: | jTLEX provides a Java implementation of TimeLine EXtraction algorithm and utilities for programmatic manipulation of TimeML graphs. |
| Outcome: | jTLEX provides a Java implementation of the TimeLine EXtraction algorithm, along with utilities for programmatic manipulation of TimeML graphs. |
A Straightforward Approach to Narratologically Grounded Character Identification (2020.coling-main)
Copied to clipboard
| Challenge: | Existing definitions of character are based on simplified or implicit definitions that do not capture essential distinctions between characters and other referents in narratives. |
| Approach: | They propose a narratologically grounded definition of character that is based on clear narrological principles and annotated 170 narrative texts. |
| Outcome: | The proposed definition of character is based on clear narratological principles and can be reliably annotated (0.78 Cohen’s ). |
GOLEM: GOld Standard for Learning and Evaluation of Motifs (2024.lrec-main)
Copied to clipboard
| Challenge: | Motifs are distinctive, recurring, widely used idiom-like words or phrases, often originating from folklore, whose meaning are anchored in a narrative. |
| Approach: | They present a dataset annotated for motific information in English . it contains 26,078 motif candidates across 34 motif types from three cultural or national groups: Jewish, Irish, and Puerto Rican. |
| Outcome: | The first dataset annotated for motific information identifies 26,078 motif candidates across 34 motif types from three cultural or national groups: Jewish, Irish, and Puerto Rican. |
Holistic Evaluation of Automatic TimeML Annotators (2022.lrec-1)
Copied to clipboard
| Challenge: | TimeML is an annotation scheme for representing temporal information in texts. |
| Approach: | They propose to combine eight metrics for holistic evaluation of TimeML graphs. |
| Outcome: | The proposed system produces graphs with 1/3 of the time indeterminacy and 1/3 of gold standard . the proposed system is compared with four other systems and is a good fit for the proposed task. |
pyTLEX: A Python Library for TimeLine EXtraction (2024.eacl-demo)
Copied to clipboard
| Challenge: | TimeML is a markup language for temporal information in text. |
| Approach: | pyTLEX is an implementation of the TimeLine EXtraction algorithm . it allows users to parse TimeML annotations, construct TimeML graphs, and execute the algorithm based on TimeML . |
| Outcome: | pyTLEX is an implementation of the TimeLine EXtraction algorithm . it allows users to parse TimeML annotations, construct TimeML graphs, and execute the algorithm to effect complete timeline extraction. |
Inducing Stereotypical Character Roles from Plot Structure (2021.emnlp-main)
Copied to clipboard
| Challenge: | Stereotypical character roles are important aids to narrative understanding and are often referred to as archetypes or dramatis personae. |
| Approach: | They propose an unsupervised method for learning stereotypical roles given only structural plot information using Vladimir Propp’s structural theory of Russian folktales. |
| Outcome: | The proposed method induces six out of seven of Vladimir Propp’s dramatis personae with F1 measures of up to 0.70 (0.58 average), with an additional category for minor characters. |
Evaluating Information Loss in Temporal Dependency Trees (2020.lrec-1)
Copied to clipboard
| Challenge: | Temporal Dependency Trees (TDTs) are an alternative to full temporal graphs for representing the temporal structure of texts. |
| Approach: | They propose a method to quantify temporal indeterminacy using temporal constraint problems to extract timelines from temporal graphs. |
| Outcome: | The proposed method shows that the tree form of TDTs results in a 109% increase in temporal indeterminacy over their corresponding temporal graphs. |
A New Approach to Animacy Detection (C18-1)
Copied to clipboard
| Challenge: | Animacy is a property for a referent to be an agent, and prior work has classified words as either animate or inanimate. |
| Approach: | They propose a method that uses supervised machine learning and hand-built rules to classify the animacy of co-reference chains. |
| Outcome: | The proposed method achieves state-of-the-art performance on a 142-text dataset . it leverages word embeddings over referring expressions, parts of speech, and grammatical and semantic roles . |