AbstractTime and again we are faced, in a number of collaborative research projects, with the challenge of interconnecting various language processing tools to implement certain industry-driven use cases focusing, for the most part, upon digital content curation processes. In this paper we first describe several of the relevant projects and their technology platforms, followed by a description of the corresponding use cases and their requirements. The content curation platform we focus upon in this article and which has been implemented as a prototype makes use of a large number of NLP services, which we also build upon for other use cases and prototypes. In addition to the implemented NLP services, the article presents a workflow manager for the flexible creation and customisation of processing workflows that make use of the above mentioned NLP services. Based on the four key principles of generality, flexibility, scalability and efficiency, we present the first version of the workflow manager by providing details on its custom definition language, explaining the communication components and the general system architecture and setup. The paper also addresses challenges in interoperability across different NLP tasks and hardware-based resource use.

AbstractConversations (normal speech) or professional interactions (e.g. projected speech in the classroom) have been identified as situations which increase individuals’ risk of exposure to respiratory viruses (including SARS-CoV-2) due to the high production of potentially infectious droplets. The few studies addressing this topic contain several methodological and linguistic limitations. This paper describes and validates an original combination of various methods, aimed at providing a global understanding of the complex physiological mechanisms underlying droplet emission during speech production. Twenty-one French speakers produced pseudowords and sentences under different phonetic conditions (e.g. loud vs normal intensity). In Experiment 1, we measured the airflow volume and airflow velocity exhaled from the mouth during speech. In Experiment 2, we measured the airflow velocity exhaled from different positions in the space around the mouth. In Experiment 3, we measured the number and size of expelled droplets. In all experiments, participants were asked to produce pseudowords and sentences under different phonetic conditions in an interactive setting. To validate our methodology, we tested the impact of voice intensity on the physiological measurements. We found that pseudowords and sentences spoken with loud intensity generated increased airflow volume and velocity compared to those spoken with normal intensity. Additionally, the number of droplets was higher for pseudowords spoken with loud intensity compared to normal intensity. From a methodological point of view, our study went beyond previous research by using multiple measures characterising droplet emission during speech. Furthermore, we applied an innovative experimental design, considering droplet emission in an interactive linguistic setting.

AbstractWeb-based platforms offer suitable experimental environments enabling the construction and reuse of natural language processing (NLP) pipelines. However, systematic evaluation of NLP tools in an open science web-based setting is still a challenge, as suitable experimental environments for the construction and reuse of NLP pipelines are still rare. This paper presents TextFlows, an open-source web-based platform, which enables user-friendly construction, sharing, execution, and reuse of NLP pipelines. It demonstrates that TextFlows can be easily used for systematic evaluation of new NLP components by integrating seven publicly available open-source part of speech (POS) taggers from popular NLP libraries, and evaluating them on six annotated corpora. The integration of new tools into TextFlows supports tools reuse, while the use of precomposed algorithm comparison and evaluation workflows supports experiment reproducibility and testing of future algorithms in the same experimental environment. Finally, to showcase the variety of evaluation possibilities offered in the TextFlows platform, the influence of various factors, such as the training corpus length and the use of pre-trained models, have been tested.