Abstract Meaning Representation (AMR) is a semantic representation that can enhance natural language generation (NLG) by providing a logical semantic input. In this paper, we propose the AMR-TST, an AMR-based text style transfer (TST) technique. The AMR-TST converts the source text to an AMR graph and generates the transferred text based on the AMR graph modified by a TST policy named style rewriting. Our method combines both the explainability and diversity of explicit and implicit TST methods. The experiments show that the proposed method achieves state-of-the-art results compared with other baseline models in automatic and human evaluations. The generated transferred text in qualitative evaluation proves the AMR-TST have significant advantages in keeping semantic features and reducing hallucinations. To the best of our knowledge, this work is the first to apply the AMR method focusing on node-level features to the TST task.

<iframe src="https://app.sli.do/event/8PaBoicsuHK7beXFunvrtr/embed/polls/d5dde25a-f42b-45da-b7b7-d5cff09c8dcf" width="300" height="400"></iframe>

AMR-TST: Abstract Meaning Representation-based Text Style Transfer

Extracting structured and grounded fact triples from raw text is a fundamental task in Information Extraction (IE). Existing IE datasets are typically collected from Wikipedia articles, using hyperlinks to link entities to the Wikidata knowledge base. However, models trained only on Wikipedia have limitations when applied to web domains, which often contain noisy text or text that does not have any factual information. We present WebIE, the first large-scale, entity-linked closed IE dataset consisting of 1.6M sentences automatically collected from the English Common Crawl corpus. WebIE also includes negative examples, i.e. sentences without fact triples, to better reflect the data on the web. We annotate ~25K triples from WebIE through crowdsourcing and introduce mWebIE, a translation of the annotated set in four other languages: French, Spanish, Portuguese, and Hindi. We evaluate the in-domain, out-of-domain, and zero-shot cross-lingual performance of generative IE models and find models trained on WebIE show better generalisability. We also propose three training strategies that use entity linking as an auxiliary task. Our experiments show that adding Entity-Linking objectives improves the faithfulness of our generative IE models.

<iframe src="https://app.sli.do/event/6qSnTPZvioRLfFJpZUjhCw/embed/polls/f51c1033-4ec9-40fd-81f3-473d189153b7" width="300" height="400"></iframe>

WebIE: Faithful and Robust Information Extraction on the Web

One vital job behavior that hospitality organizations should support is helping behavior, as it represents employees’ proactive and reactive responses to the needs of their colleagues (Hwang et al., 2021) and plays a critical role in various service aspects (e.g., Zou et al., 2015; Kim et al., 2018; Özduran and Tanova, 2017). However, there is a paucity of research on how to support employees to engage in helping behavior. Thus, the aim of this study is to increase the understanding of how an employee’s living a calling and cooperative organizational culture impact their workplace well-being, and ultimately their helping behavior toward coworkers. The sample was collected from US hospitality employees using an MTurk panel. Partial Least Squares structural equation modeling and multigroup analysis were used to explain the proposed model. Results revealed that living a calling and cooperative organizational culture positively influenced both workplace well-being and helping behavior, respectively, and workplace well-being positively influenced helping behavior. Additionally, cooperative organizational culture and workplace well-being were significant mediators between living a calling and helping behavior. Cooperative organizational culture was vital for female employees to engage in helping behavior, while the impact of workplace well-being on helping behavior was more important for male employees. This study advances the understanding of employee’s helping behavior by unveiling new motivators (i.e., living a calling, cooperative organization culture, workplace well-being) and extends the understanding of organizational support theory and social identity theory. The study also demonstrates that gender has a significant influence on helping behaviors. Hospitality managers will benefit from this research by understanding the importance of employee’s social identifications and the necessary supportive workplace environment to elevate employee’s well-being and their helping behavior toward coworkers, a critical aspect to increase competitive advantage for organizational success.

The effect of cooperative organizational culture and gender on helping behavior

Machine Learning

Artificial Intelligence is already a part of humans everyday life activities, and we are just at the beginning.
Underline presents the up-to-date research in all fields of AI.

Welcome to Etsy's knowledge hub!

Featured presentations

Content from leading global institutions

Machine Learning

Artificial Intelligence

Business, Management and Economics

Customer Care

Stay up to date with the latest Underline news!

PRODUCT

COMPANY

RESOURCES