Contents

Properties

Use this page to Add, Edit or Delete speech recognition properties. These properties constitute name and value pairs that are maintained by the recognition subsystem. They are retrieved and modified through the ‘Reco Get Property’ and ‘Reco Set Property’ tools.

The “namespace prefix” classifies properties. This is important to distinguish between VoiceXML, I3, and engine specific properties. Properties that have no namespace prefix correspond to their VoiceXML counterparts of the same name. See the table below for a list of property names and values.

Note: For more information on the tools, see the Handler help in the PureConnect Documentation Library.

To add a property click Add and enter the Property Name. Click Add Value and enter the property value. Click Remove Value to delete a property.

To edit a property select the property and click Edit. Make changes then click OK to save your changes.

To delete a property select the property and click Delete.

Property Name	Description and Value
reco:ASREngine	This is the name of the ASR (Automatic Speech Recognition) engine used by the current session.
reco:ASRServer	Machine name of the ASR server on which the port of the session is located. Empty string if lazy port allocation and port has not yet been allocated.
reco:ASRSupportedFeatures	This Property is reserved for future release. Space separated list of features supported by the ASR engine of a session. Handlers can enforce support for features in the ‘Reco Initialize’ and check for individual support through the ‘Reco Has Feature’ tool. The following features are currently defined: asr_multiple_grammars - The ASR engine supports multiple simultaneously active grammars (multiple ASR grammars may be referenced by the ‘Reco Input’ tool at once). asr_grammar_weights - The ASR engine supports weights for individual grammars. If not supported, 1.0 is assumed. asr_grammar_fragments - The ASR engine supports fragments in the grammar URIs. asr_sisr_conditionals - The engine supports conditional expressions in semantic interpretation tags. asr_sisr_func_reject -The engine supports the ‘Reject’ function in semantic interpretation tags. asr_inline_grammars - The ASR engine supports adding grammars defined as source at runtime, thus the tools ‘Reco Register Grammar String’ and ‘Reco Register Inline Grammar’ are supported. asr_agile - Engine does not tie the session.
reco:ASRMinResultConf	Minimum confidence an ASR recognition hypothesis must have in order to be included in the recognition result. Default: "0.1"
reco:GrammarBaseURI	Base URI used to resolve relative grammar URIs.
reco:ASRDebugWaveLogging	Boolean property to enable logging of wave data for debugging purposes. The default is "false." Note: When the reco:RestrictResultTracing property is set to True, the reco:ASRDebugWaveLogging property will automatically be set to False.
reco:RestrictResultTracing	The purpose of this property is to prevent sensitive data from being logged in the IP logs. Set this property to True for an input operation. This sets a flag in the recognition result that suppresses tracing of the utterance and slot values in the trace logs. Note: When this property is set to True, it will also disable the ASR Debug Wav recordings.
reco:ASRBargeinDisabled	Boolean property to disable barge-in. If this property is set to "true", the plays are not cancelled when speech is detected. The plays will be cancelled when the recognition is complete (successful or failed). Default: "false"
reco:ASRANII	Defines the ANII value of the current call. Used for logging.
reco:ASRApplicationName	Defines the Application Name to be logged for the current call.
reco:ASRConfidenceMapping	Defines a list of confidence mapping pairs which are used to map the native engine confidences into the confidence values returned in the recognition result. This property is thus used to normalize differing confidence values returned by the ASR engines. Each mapping pair consists of two floating point values between 0.0 and 1.0, separated by an equal sign. The first value designates the native value (returned by the engine) and the second value the resulting mapped confidence for that particular engine value. The property may contain an arbitrary number of mapping points. Linear interpolation is used to map confidence values that lie between two point. When mapping confidence calues, two implicit mapping points "0.0=0.0" and "1.0=1.0" are always assumed. This results in an identity mapping if no propery is defined and to ensure correct handling at the boundaries.
reco:EnableSpeakerVerification	Specifies whether speaker identification/verification features should be enabled for an engine.
reco:GrammarCompileTimeout	Specifies the timeout (in seconds) to use when waiting on grammars to be compiled. Depending on the engine, this may not be able to be set through IA and may only have an affect if set in the server's xml config file. Default: 300.0 (5 minutes)
reco:EndOfSpeechNotifyDelay	Specifies the delay after the end of speech has been detected before the server sends an end-of-speech notification. Default: 0.1 (100 ms)
recocfg:ASRMaxProxySessionCount	Maximum number of sessions a server proxy may host. Creating more sessions than the maximum number will be refused. Sessions will not be killed if the maximum session count is changed to a lower value than the number of currently active sessions.
recocfg:ASRServerProxyEnabled	Configuration property specifying whether a server proxy is enabled. A disabled proxy will not allow new sessions (similar proxies whose recocfg:MaxProxySessionCount property are set to 0.
recocfg:AudioAdvertisedAddress	IP address or DNS name under which the CIC server sees the RTP endpoint of the ASR server. The address specified through this property must refer to the same physical device as recocfg:AudioListenInterfaceAddress. IMPORTANT: This property must be used with care and only if the implications are well understood. Misconfiguration may lead to hard to diagnose problems or failures in the audio delivery to the ASR servers.
recocfg:AudioConnectionProbeInterval	Interval between audio connection probes in seconds. It specifies the minimum interval between successful connection probe attempts.
recocfg:AudioConnectionProbeMaxFail	Maximum number of failed audio connection probe attempts before a connection configuration problem is assumed and reported.
recocfg:AudioListenInterfaceAddress	Local IP address of the NIC which the server should use to receive audio. This address must refer to the same physical device as the recocfg:AudioAdvertisedAddress property. The local NIC address must be specified as "dot" address (no DNS name). IMPORTANT: This property must be used with care and only if the implications are well understood. Misconfiguration may lead to hard to diagnose problems or failures in the audio delivery to the ASR servers.
recocfg:EnableAudioConnectionProbe	Enables the audio connection probes to check the connection between the CIC servers and ASR servers at regular intervals.
recocfg:ServerProxyPriority	Priority of a server proxy compared to other server proxies for the same engine. When choosing a server for a session, the proxies with low priorities are picked first until their session limit is reached.
recocfg:SupportedLanguages	List of languages supported by an EIM or ASR server. Note: A value of "en" indicates that all versions of English are supported. A value of "en-US" indicates that only the U.S. English version is supported.
sensitivity	VoiceXML related property: Relative sensitivity to speech input. 1.0 means highly sensitive to quiet input and 0.0 very insensitive to noise. Not all engines will support this. Default: 0.5
speedvsaccuracy	VoiceXML related property: Relative tradeoff between recognition accuracy and CPU utilization (faster recognition). 0.0 means fastest recognition, 1.0 means highest recognition accuracy. Not all engines support this. Default: 0.5
bargeintype	Type of barge-in performed by the voice input. This property corresponds to the ‘bargeintype’ VoiceXML property. Possible values defined by VoiceXML: speech - The barge-in occurs when speech or DTMF is detected. hotword - The barge-in occurs after a grammar has accepted voice or DTMF input. Input that does not match a grammar is ignored. Thus, the tool will never take the “No Match” exit. The default is “speech”.