Title: Story Fragment Stitching: The Case of the Story of Moses
We introduce the task ofstory fragment stitching,which is the process of automatically aligning andmerging event sequences of partial tellings of astory (i.e.,story fragments). We assume that eachfragment contains at least one event from the storyof interest, and that every fragment shares at leastone event with another fragment. We propose agraph-based unsupervised approach to solving thisproblem in which events mentions are representedas nodes in the graph, and the graph is compressedusing a variant of model merging to combine nodes.The goal is for each node in the final graph to con-tain only coreferent event mentions. To find coref-erent events, we use BERT contextualized embed-ding in conjunction with atf-idfvector representa-tion. Constraints on the merge compression pre-serve the overall timeline of the story, and the finalgraph represents the full story timeline. We evalu-ate our approach using a new annotated corpus ofthe partial tellings of the story of Moses found inthe Quran, which we release for public use. Ourapproach achieves a performance of 0.63F1score more »« less
A crucial step in the construction of any event story or news report is to identify entities involved in the story, such entities can come from a larger background knowledge graph or from a text corpus with entity links. Along with recognizing which entities are relevant to the story, it is also important to select entities that are relevant to all aspects of the story. In this work, we model and study different types of links between the entities with the goal of identifying which link type is most useful for the entity retrieval task. Our approach demonstrates the e
Story Programming is an approach for teaching complex computational and algorithmic thinking skills using simple stories anyone can relate to. One could learn these skills independent of a computer or with the use of a computer as a tool to interact with the computation in the tale. This research study examines the use of Story Programming before teaching coding in a computer science orientation course to determine if it is a viable alternative to the code-focused way of teaching the class in the past. We measure the viability of the Story Programming approach by evaluating student-success and learning outcomes, as well as student reactions to post-survey questions.
Proc. 2023 ACM SIGIR Int. Conf. on Research and Development in Information Retrieval
(Ed.)
Unsupervised discovery of stories with correlated news articles in real-time helps people digest massive news streams without expensive human annotations. A common approach of the existing studies for unsupervised online story discovery is to represent news articles with symbolic- or graph-based embedding and incrementally cluster them into stories. Recent large language models are expected to improve the embedding further, but a straightforward adoption of the models by indiscriminately encoding all information in articles is ineffective to deal with text-rich and evolving news streams. In this work, we propose a novel thematic embedding with an off-the-shelf pretrained sentence encoder to dynamically represent articles and stories by considering their shared temporal themes. To realize the idea for unsupervised online story discovery, a scalable framework USTORY is introduced with two main techniques, theme- and time-aware dynamic embedding and novelty aware adaptive clustering, fueled by lightweight story summaries. A thorough evaluation with real news data sets demonstrates that USTORY achieves higher story discovery performances than baselines while being robust and scalable to various streaming settings.
Sorosh, Shokrullah; Zhang, Jiachen; Hutchinson, Tara; Ryan, Keri; Smith, Kevin; Kovac, Adam; Pei, Shiling; Barbosa, Andre; Simpson, Barbara
(, 18th U.S.-Japan-New Zealand Workshop on the Improvement of Structural Engineering and Resilience)
A series of shake table tests were recently conducted on full-scale 10-story and 6-story mass timber buildings at the 6-DOF Large High-Performance Outdoor Shaking Table facility at the University of California San Diego. Stairs, providing the primary egress in and out of a building during and after an earthquake event, were incorporated in each of these building test programs. To ensure they support the immediate recovery of building function, a variety of drift-release details were incorporated. Previous earthquake events and experimental studies have shown that stairs are among the most drift-sensitive nonstructural systems and are prone to damage, therefore relieving interstory drifts is paramount to improving their performance. To this end, the designed drift-release connections within the stairs considered the test buildings response during earthquake motions scaled at various hazard levels with expected minor and repairable damage under large earthquake loading. This paper provides an overview of the shake table test programs from the perspective of the design and performance of resilient steel stairs.
Zhang, Yunyi; Guo, Fang; Shen, Jiaming; Han, Jiawei
(, KDD'22:The 28th {ACM} {SIGKDD} Conference on Knowledge Discovery and Data Mining, August 14-18, 2021)
Automated event detection from news corpora is a crucial task towards mining fast-evolving structured knowledge. As real-world events have different granularities, from the top-level themes to key events and then to event mentions corresponding to concrete actions, there are generally two lines of research: (1) theme detection tries to identify from a news corpus major themes (e.g., “2019 Hong Kong Protests” versus “2020 U.S. Presidential Election”) which have very distinct semantics; and (2) action extraction aims to extract from a single document mention-level actions (e.g., “the police hit the left arm of the protester”) that are often too fine-grained for comprehending the real-world event. In this paper, we propose a new task, key event detection at the intermediate level, which aims to detect from a news corpus key events (e.g., HK Airport Protest on Aug. 12-14), each happening at a particular time/location and focusing on the same topic. This task can bridge event understanding and structuring and is inherently challenging because of (1) the thematic and temporal closeness of different key events and (2) the scarcity of labeled data due to the fast-evolving nature of news articles. To address these challenges, we develop an unsupervised key event detection framework, EvMine, that (1) extracts temporally frequent peak phrases using a novel ttf-itf score, (2) merges peak phrases into event-indicative feature sets by detecting communities from our designed peak phrase graph that captures document cooccurrences, semantic similarities, and temporal closeness signals, and (3) iteratively retrieves documents related to each key event by training a classifier with automatically generated pseudo labels from the event-indicative feature sets and refining the detected key events using the retrieved documents in each iteration. Extensive experiments and case studies show EvMine outperforms all the baseline methods and its ablations on two real-world news corpora.
Aldawsari, Mohammed, Asgari, Ehsaneddin, and Finlayson, Mark A. Story Fragment Stitching: The Case of the Story of Moses. Retrieved from https://par.nsf.gov/biblio/10220128. 1st Workshop on Artificial Intelligence for Narratives (AI4N 2020) .
Aldawsari, Mohammed, Asgari, Ehsaneddin, & Finlayson, Mark A. Story Fragment Stitching: The Case of the Story of Moses. 1st Workshop on Artificial Intelligence for Narratives (AI4N 2020), (). Retrieved from https://par.nsf.gov/biblio/10220128.
Aldawsari, Mohammed, Asgari, Ehsaneddin, and Finlayson, Mark A.
"Story Fragment Stitching: The Case of the Story of Moses". 1st Workshop on Artificial Intelligence for Narratives (AI4N 2020) (). Country unknown/Code not available. https://par.nsf.gov/biblio/10220128.
@article{osti_10220128,
place = {Country unknown/Code not available},
title = {Story Fragment Stitching: The Case of the Story of Moses},
url = {https://par.nsf.gov/biblio/10220128},
abstractNote = {We introduce the task ofstory fragment stitching,which is the process of automatically aligning andmerging event sequences of partial tellings of astory (i.e.,story fragments). We assume that eachfragment contains at least one event from the storyof interest, and that every fragment shares at leastone event with another fragment. We propose agraph-based unsupervised approach to solving thisproblem in which events mentions are representedas nodes in the graph, and the graph is compressedusing a variant of model merging to combine nodes.The goal is for each node in the final graph to con-tain only coreferent event mentions. To find coref-erent events, we use BERT contextualized embed-ding in conjunction with atf-idfvector representa-tion. Constraints on the merge compression pre-serve the overall timeline of the story, and the finalgraph represents the full story timeline. We evalu-ate our approach using a new annotated corpus ofthe partial tellings of the story of Moses found inthe Quran, which we release for public use. Ourapproach achieves a performance of 0.63F1score},
journal = {1st Workshop on Artificial Intelligence for Narratives (AI4N 2020)},
author = {Aldawsari, Mohammed and Asgari, Ehsaneddin and Finlayson, Mark A.},
editor = {null}
}
Warning: Leaving National Science Foundation Website
You are now leaving the National Science Foundation website to go to a non-government website.
Website:
NSF takes no responsibility for and exercises no control over the views expressed or the accuracy of
the information contained on this site. Also be aware that NSF's privacy policy does not apply to this site.
An official website of the United States government
