BAS Speech Science Web Services - an Update of Current Developments

BAS Speech Science Web Services – an Update on Current Developments Thomas Kisler∗, Uwe D. Reichel†, Florian Schiel∗, Christoph Draxler∗, Bernhard Jackl∗, Nina P¨orner∗ ∗Bavarian Archive for Speech Signals, Ludwig-Maximilians-Universit¨at M¨unchen Schellingstr. 3, 80799 M¨unchen, Germany {kisler,schiel,draxler,jackl,poerner}@bas.uni-muenchen.de †Research Institute for Linguistics, Hungarian Academy of Sciences Bencz´ur u. 33, 1068 Budapest, Hungary uwe.reichel@nytud.mta.hu Abstract In 2012 the Bavarian Archive for Speech Signals started providing some of its tools from the field of spoken language in the form of Software as a Service (SaaS). This means users access the processing functionality over a web browser and therefore do not have to install complex software packages on a local computer. Amongst others, these tools include segmentation & labeling, grapheme-to-pho- neme conversion, text alignment, syllabification and metadata generation, where all but the last are available for a variety of languages. Since its creation the number of available services and the web interface have changed considerably. We give an overview and a detailed description of the system architecture, the available web services and their functionality. Furthermore, we show how the number of files processed over the system developed in the last four years. Keywords: automatic segmentation, grapheme-to-phoneme, syllabification, forced alignment, RESTful web service, web interface 1. Introduction With the emergence of standards for modern web technolo- gies and, at least equally important, browsers that support them, it is possible to implement even complex software as web applications that run in a browser. Since most users already have a modern browser installed on their computer, there is no need for additional software. In late 2014 the new HTML5 standard has been approved and is now sup- ported acceptably by most modern browsers (Hickson et al., 2014; Leenheer, 2015). Software-as-a-service (SaaS) means that software is no longer purchased once by the cus- tomer, but rather payed for when accessed. In the academic world the core concept stays the same, even if software is often offered for free. This means that the user accesses a software resource on-demand running on the infrastructure of the software provider, without the need to install spe- cialized software on her computer (Buxmann et al., 2008). Some examples for SaaS, next to ours, are the EmuLabeller, a tool for online labeling of speech signals (Winkelmann and Raess, 2014), WebLicht, an engine that enables the user to construct linguistic chains to automatically anno- tate text corpora (Hinrichs et al., 2010) and WebAnno, a tool for doing manual text annotation in the browser (de Castilho et al., 2014). For other available resources within the CLARIN infrastructure, please check CLARIN (2015). The new HTML5 standard, as Anthes (2012) states, does not only have advantages for users, but also for develop- ers. With modern browsers supporting many of the spec- ified technologies and state-of-the-art debuggers, develop- ing software for the web has become much easier. In the last four years, the Bavarian Archive for Speech Signals (BAS) transformed a few of its most used appli- cations to services. As these tools have historically been Linux command line …

Content truncated to 500 words