New Corpora

2013 Data Pack: create a custom collection of eight corpora from among LDC’s 2013 publications, available to not-for-profit and government organizations through September 15, 2015 only

Phonetic Segmentation and Tone Labels: Mandarin Chinese Phonetic Segmentation and Tone:7,849 Mandarin Chinese utterances derived from broadcast recordings, segmented and labeled and separated into training and test sets

Subglottal Resonances: The Subglottal Resonances Database:developed by Washington University and University of California Los Angeles, 45 hours of simultaneous microphone and subglottal accelerometer recordings of 25 adult male and 25 adult female speakers of American English for use in automatic speech recognition and in studies of speech production, perception and technology 

Code Switching Speech: Mandarin-English Code-Switching in South-East Asia: developed by Nanyang Technological University and Universiti Sains Malaysia, 192 hours of Mandarin-English code-switching speech from 156 speakers with associated transcripts 

Email Communications: Avocado Research Email Collection: 279 email accounts from a defunct technology company inclusive of emails, attachments and processed personal folders with metadata describing folder structure, email characteristics, calendar entries and contacts for use in social network analysis, e-discovery and related fields

Multichannel Noisy Telephone Speech: RATS Speech Activity Detection: developed by LDC, 3,000 hours of multilingual conversational telephone speech passed through eight distinct transceiver configurations with severely degraded audio signals typical of various radio communication channels

GALE Parallel, Word Aligned and Tagged Text:  Arabic/Chinese and English parallel, word aligned and tagged resources LDC developed for the DARPA GALE program

GALE Phase 3 and 4 Arabic Broadcast News Parallel Text: 86 source-translation document pairs comprising 325,538 words of Arabic source text and its English translation

GALE Phase 3 and 4 Arabic Broadcast Conversation Parallel Text: 55 source-translation document pairs comprising 280,535 words of Arabic source text and its English translation

GALE Chinese-English Parallel Aligned Treebank -- Training: 229,249 tokens of word aligned Chinese and English parallel text with treebank annotations

GALE Chinese-English Word Alignment and Tagging -- Broadcast Training Part 3: 242,020 tokens of word aligned Chinese-English parallel text enriched with linguistic tags.