Search for: All records

Large-scale, high-quality corpora are critical for advancing research in coreference resolution. However, existing datasets vary in their definition of coreferences and have been collected via complex and lengthy guidelines that are curated for linguistic experts. These concerns have sparked a growing interest among researchers to curate a unified set of guidelines suitable for annotators with various backgrounds. In this work, we develop a crowdsourcing-friendly coreference annotation methodology, ezCoref, consisting of an annotation tool and an interactive tutorial. We use ezCoref to re-annotate 240 passages from seven existing English coreference datasets (spanning fiction, news, and multiple other domains) while teaching annotators only cases that are treated similarly across these datasets. Surprisingly, we find that reasonable quality annotations were already achievable (90% agreement between the crowd and expert annotations) even without extensive training. On carefully analyzing the remaining disagreements, we identify the presence of linguistic cases that our annotators unanimously agree upon but lack unified treatments (e.g., generic pronouns, appositives) in existing datasets. We propose the research community should revisit these phenomena when curating future unified annotation guidelines. 

https://doi.org/10.7275/zdg0-0914 

The study of language variation examines how language varies between and within different groups of speakers, shedding light on how we use language to construct identities and how social contexts affect language use. A common method is to identify instances of a certain linguistic feature - say, the zero copula construction - in a corpus, and analyze the feature’s distribution across speakers, topics, and other variables, to either gain a qualitative understanding of the feature’s function or systematically measure variation. In this paper, we explore the challenging task of automatic morphosyntactic feature detection in low-resource English varieties. We present a human-in-the-loop approach to generate and filter effective contrast sets via corpus-guided edits. We show that our approach improves feature detection for both Indian English and African American English, demonstrate how it can assist linguistic research, and release our fine-tuned models for use by other researchers. 

https://aclanthology.org/2022.nlpcss-1.11 

Greenhouses conserve land and water while increasing crop production, making them an attractive system for low environmental impact agriculture. Yet, to achieve this goal, there is a need to reduce their large energy demand. Employing semitransparent organic solar cells (OSCs) on greenhouse structures provide an opportunity to offset the greenhouse energy needs while maintaining the lighting needs of the plants. However, the design trade-off involved in optimizing solar power generation and crop productivity to maximize greenhouse economic value is yet to be studied in detail. Here, a functional plant growth model is integrated with a dynamic energy model that includes supplemental lighting to optimize the economics of growing lettuce and tomato. The greenhouse optimization considers 64 different OSC active layers with varying roof coverage for 25 distinct climates providing a global perspective. We find that crop yield is the primary economic driver, and that crop yield can be maintained in OSC-greenhouses across diverse climates. The crop productivity along with the energy produced by the OSCs results in improved net present value of the OSC-greenhouses relative to conventional systems in most climates for both lettuce and tomato. In addition, we find common solar cell active layers that maximize greenhouse economic value resulting in guidelines for scaling up OSC-greenhouse design. Through this model framework, we highlight the opportunity for OSCs in greenhouses, uncover designs and locations that provide the most value, and provide a basis for further development of OSC-greenhouses to achieve a sustainable means of food production.