Pre-Grant Publication Number: 20110112833
Please help the USPTO examine the application by evaluating the relevance of the publicly submitted prior art to the patent application.
Peer To Patent forwards the Top 10 most relevant prior art submissions and their annotations to the USPTO.
Review this prior art and click on the thumbs up (or down) to indicate whether this submission should be forwarded to the USPTO.
If you login then you can add an annotation by typing in the box at the bottom of the screen to comment on the relevance of the prior art to the claims of the patent application.
Review this prior art and click on the thumbs up (or down) to indicate whether this submission should be forwarded to the USPTO.
If you login then you can add an annotation by typing in the box at the bottom of the screen to comment on the relevance of the prior art to the claims of the patent application.
Prior Art Detail
Summary / Description
| Summary / Description | We describe a system for automating call-center analysis and monitoring. Our system integrates transcription of incoming calls with analysis of their content; for the analysis, we introduce a novel method of estimating the domain-specific importance of conversation fragments, based on divergence of corpus statistics. |
Basic Information
| Type of Prior Art | Print Publication |
| Publication Title * | Automatic Analysis of Callcenter Conversations |
| Author | Gilad Mishne et al. |
| ISBN | |
| Page Range | |
| Medium | Journal article |
| Publication Date * | January 1, 2005 |
| URL | |
Notes / To Do
| Notes | |
Excerpt
Excerpt The transcription server, used for transcribing the callcenter
data, is an IBM research prototype; it was built on
top of existing IBM core WebSphere technology, in particular
WebSphere Voice Server [14] and its major components
such as the speech recognition engine facilities. The transcription
server is able to get an audio stream or a URI to a
prerecorded audio file, and transcribe it. The transcription
output is an XML file, which includes the transcript and
some meta-data; for example, each word has time-stamps
of its beginning and its end. These time-stamps enable the
synchronization of the audio and the text data. The server
uses the latest Large Vocabulary Continuous Speech Recognition
(LVCSR) technology developed in IBM research [8],
to transcribe continuous spontaneous 6KHz speech recorded
by a call-center into text. The output words are selected
from a large US English vocabulary, which has a good coverage
of the spoken language. In general, an LVCSR system
includes an engine, a vocabulary, a language model and an
acoustic model. Those define the scope of transcription,
the language domain in which it is expected to be, and
the acoustic environment of the recordings. The full process
can be characterized by several steps: the input speech
signal is first analyzed into a sequence of acoustic feature
vectors, and then statistical methods are used to determine
the most probable word sequence. In the latter stage, the
acoustic model is used for computing the probability of the
observed sequence of vectors given a word sequence and the
language model for computing the probability of a word sequence,
independently of the speech input. The recognition
process is carried out by maximizing the product of the two
(Bayes law). |
Relevance
Claims
1
Relevance
This journal article from 2005 describes a text-to-speech systems for a call center. This same technology, applied multiple times, could make this patent obvious.
This journal article from 2005 describes a text-to-speech systems for a call center. This same technology, applied multiple times, could make this patent obvious.
Claim Chart
Some
0 days left
