In this paper we address the problem of extracting the identities and phone numbers of the callers in voicemail messages. Previous work in information extraction from speech includes spoken document retrieval and named entity detection. This task differs from the named entity task in that the information we are interested in is a subset of the named entities in the message, and consequently, the need to pick the correct subset makes the problem more difficult. Also, the caller's identity may include information that is not typically associated with a named entity. In this work, we present two information extraction methods, one based on hand-crafted rules, and one based on a maximum entropy model. We find that both systems give good performance when applied to manually-derived transcriptions, and that the maximum entropy system can reliably identify the time intervals containing phone numbers, even in the presence of significant decoding errors.
展开▼
