LREC 2000 2nd International Conference on Language Resources & Evaluation
 

Previous Paper   Next Paper

Title Models of Russian Text/Speech Interactive Databases for Supporting of Scientific, Practical and Cultural Researches
Authors Skrelin Pavel (Department of Phonetics, St. Petersburg State University, Universitetskaja nab. 11, 199034, Russia, paul@phonet.lang.pu.ru)
Sherstinova Tatiana (Department of Phonetics, St. Petersburg State University, Universitetskaja nab. 11, 199034, Russia, tanya@ts4306.spb.edu)
Keywords Databases, Internet, Phonetics, Scientific and Cultural Research, Speech
Session Session SP5 - Multimodal - Multimedia Resources and Tools
Full Paper 185.ps, 185.pdf
Abstract The paper briefly describes the following databases: ”Online Sound Archives from St. Petersburg Collections”, ”Regional Variants of the Russian Speech”, and ”Multimedia Dictionaries of the minor Languages of Russia”, the principle feature of which is the built-in support for scientific, practical and cultural researches. Though these databases are addressed to researchers engaged mainly in Spoken Language Processing and because of that their main object is Sound, proposed database ideology and general approach to text/speech data representation and access may be further used for elaboration of various language resources containing text, audio and video data. Such approach requests for special representation of the database material. Thus, all text and sound files should be accompanied by information on their multi-level segmentation, which should allow the user to extract and analyze any segment of text or speech. Each significant segment of the database should be perceived as a potential object of investigation and should be supplied by tables of descriptive parameters, mirroring its various characteristics. The list of these parameters for all potential objects is open for further possible extension.