Abstract: A telecommunication system is arranged between a trusted automated interactive system and a non-secure agent. The trusted automated interactive portion is operable for bidirectional communications with a caller. The agent portion is arranged to be bridged onto a communication connection between the caller (or network chat user) and the trusted interactive system under certain conditions. Sensitive data transmitted between the system and a calling user may be blocked from being communicated to the agent even while the agent is bridged onto the connection. In one embodiment, information (such as a verification of authenticity of the caller) pertaining to the sensitive data (but not the sensitive data itself) is communicated to the agent.

Abstract: There is disclosed a system and method for automatically performing semantic categorization. In one embodiment at least one text description pertaining to a category set is accepted along with words that are anticipated to be uttered by a user pertaining to that category set; lexical chaining confidence score is attached to each pair matched between the anticipated words and the accepted text description. These confidence scores are used subsequently by a categorization circuit that accepts a text phrase utterance from an input source along with a category set pertaining to the accepted utterance. The categorization circuit, in one embodiment, creates word pairs matched between the accepted text phrase utterance and the accepted category set. From these word scores, the category pertaining to the utterance is determined based, at least in part, on the assigned lexical chaining confidence scores as previously determined.

Abstract: A telecommunication system is arranged between a trusted automated interactive system and a non-secure agent. The trusted automated interactive portion is operable for bidirectional communications with a caller. The agent portion is arranged to be bridged onto a communication connection between the caller (or network chat user) and the trusted interactive system under certain conditions. Sensitive data transmitted between the system and a calling user may be blocked from being communicated to the agent even while the agent is bridged onto the connection. In one embodiment, information (such as a verification of authenticity of the caller) pertaining to the sensitive data (but not the sensitive data itself) is communicated to the agent.

Abstract: There is disclosed a system and method for automatically performing semantic categorization. In one embodiment at least one text description pertaining to a category set is accepted along with words that are anticipated to be uttered by a user pertaining to that category set; lexical chaining confidence score is attached to each pair matched between the anticipated words and the accepted text description. These confidence scores are used subsequently by a categorization circuit that accepts a text phrase utterance from an input source along with a category set pertaining to the accepted utterance. The categorization circuit, in one embodiment, creates word pairs matched between the accepted text phrase utterance and the accepted category set. From these word scores, the category pertaining to the utterance is determined based, at least in part, on the assigned lexical chaining confidence scores as previously determined.

Abstract: A Statistical Language Model (SLM) that can be used in an ASR for Interactive Voice Response (IVR) systems in general and Natural Language Speech Applications (NLSAs) in particular can be created by first manually producing a brief description in text for each task that can be performed in an NLSA. These brief descriptions are then analyzed, in one embodiment, to generate spontaneous speech utterances based pre-filler patterns and a skeletal set of content words. The pre-filler patterns are in turn used with Part-of-Speech (POS) tagged conversations from a spontaneous speech corpus to generate a set of pre-filler phrases. The skeletal set of content words is used with an electronic lexico-semantic database and with a thesaurus-based content word extraction process to generate a more extensive list of content words. The pre-filler phrases and content words set, thus generated, are combined into utterances using a lexico-semantic resource based process.

Abstract: A system and method for a packet VRU which directly utilize packet network protocols, such as those of the H.323 standard, to provide enhanced services via a packet network, such as whisper mode communications, call recording, call broadcasting, advertisement broadcasting during calls, one-number service, and operator services. In a preferred embodiment of the present invention, the packet VRU establishes control of media streams, such as through the use of separate signaling channels associated with various communication devices, in order to provide desired enhanced call services.

Abstract: A system and method for a packet VRU which directly utilize packet network protocols, such as those of the H.323 standard, to provide enhanced services via a packet network. The packet VRU generally operates within the packet network and is not required to provide data format translation or multiple device-type access. In a preferred embodiment of the present invention, the packet VRU redirects the media stream from a source so that it is sent directly to a destination, instead of passing through the packet VRU. Alternatively, if the packet VRU must perform processing on the message contents, the packets may be sent to both the destination and to the packet VRU. The packet VRU may still retain call control over the media streams by maintaining the signaling and user input components of the call.

Abstract: A system and method are disclosed which determine the flexibility of requests and utilize such flexibility of requests in managing a finite number of resources for servicing such requests. As an exemplary implementation within a telephony call center, a caller requesting to speak with an agent of the call center may be received. The call center may ascertain whether the caller is capable and willing to speak in multiple languages, such as both English and Spanish. A caller that is so capable and willing may be more efficiently serviced by allocating the “first available” of English and Spanish speaking resources to service the caller. Thus flexible requests that are capable and willing to be serviced by any of a plurality of resources, may be utilized to more effectively manage a finite number of resources in servicing requests.

Abstract: A simplified programming language is disclosed. The main flow of the program is enclosed in one or more frames. Concurrent operations and asynchronous events that may affect the main program code are attached to the frames as decorations. Operations within the frames may loop or pause until an outside event defined in a decoration causes the main program to exit the frame or to end. Concurrently operating programs may cooperatively exchange data using the decorations. The present language enables software code to be easily updated or modified by changing the decorations that are connected to the program frames.

Abstract: A tool for scheduling the delivery of media objects within varying constraints of available bandwidth. In a preferred embodiment in which the invention is enabled by software executable on a general purpose multimedia grade computer, a system designer orchestrates the playback of a series of media objects on a playback “score” analogous to a musical score. The media objects are represented on the playback score as playback icons. Responsive to the designer's manipulation of playback icons on the playback score, the present invention further creates and displays a corresponding download score, in which download icons (corresponding to playback icons on the playback score) are deployed and arranged to fill up a displayed “pipe” of available bandwidth. The topographical area of the download icons, as displayed, corresponds to the data size of the media objects. The download icons are displayed as rectangular or square-cornered polygonal shapes.

Abstract: A system and method for administering Call Back service are disclosed which are capable of pre-qualifying such Call Back service. Such a system and method are capable of pre-qualifying a request for a Call Back service before returning a call back to the requesting customer. In this manner, the customer may not begin being charged at the lower call rate until after a connection is first established with the called number. A call back server may attempt to connect the customer requesting Call Back service with the called party by first attempting to establish a connection with the called party. Once a connection is established with the called party, the call back server may call the customer and connect the called party to the customer. In this manner, the customer may only be charged for a call back from the call back server if a connection is first established with the called number.

Abstract: A system and method for a packet Voice Response Unit (VRU) which directly utilize packet network protocols, such as those of the H.323 standard, to provide enhanced services via a packet network. The packet VRU generally operates within the packet network and is not required to provide data format translation or multiple device-type access. The packet VRU may be built entirely in software running on a network server with a standard packet network connection such as Ethernet or token-ring. In a preferred embodiment of the present invention, the packet VRU redirects the media stream from a source so that it is sent directly to a destination, instead of passing through the packet VRU. Alternatively, if the packet VRU must perform processing on the message contents, the packets may be sent to both the destination and to the packet VRU. The packet VRU may still retain call control over the media streams by maintaining the signaling and user input components of the call.

Abstract: An interval timer for timing multiple repetitive timing intervals. A single large clock register increments ticks of a high-speed clock. Successive previously-stored timing values are loaded into a single compare register which is preferably of equivalent length to the clock register. A comparator monitors the clock register's current value and compares it with the timing value currently loaded in the compare register. As the clock register's value reaches the current timing value in the compare register, an alert signal is generated and sent out to activate a particular timed operation identified by an event ID (“EID”) associated with the timing value in the compare register. The current timing value in the compare register is then discarded, and the next timing value in sequence is retrieved into the compare register. A repeat flag is carried with each timing value and associated EID.

Abstract: A system and method for a packet VRU which directly utilize packet network protocols, such as those of the H.323 standard, to provide enhanced services via a packet network. The packet VRU generally operates within the packet network and is not required to provide data format translation or multiple device-type access. In a preferred embodiment of the present invention, the packet VRU redirects the media stream from a source so that it is sent directly to a destination, instead of passing through the packet VRU. Alternatively, if the packet VRU must perform processing on the message contents, the packets may be sent to both the destination and to the packet VRU. The packet VRU may still retain call control over the media streams by maintaining the signaling and user input components of the call.

Abstract: A system and method for a packet VRU which directly utilize packet network protocols, such as those of the H.323 standard, to provide enhanced services via a packet network, such as whisper mode communications, call recording, call broadcasting, advertisement broadcasting during calls, one-number service, and operator services. In a preferred embodiment of the present invention, the packet VRU establishes control of media streams, such as through the use of separate signaling channels associated with various communication devices, in order to provide desired enhanced call services.

Abstract: A single large register increments ticks of a high-speed clock. A single compare register is associated with the clock register, the compare register preferably being of equivalent length to the clock register. Successive previously-stored timing values are then loaded into the compare register. Typically the timing values are pre-sorted in chronological order. A comparator monitors the clock register's current value and compares it with the timing value currently loaded in the compare register. As the clock register's value reaches the current timing value in the compare register, an alert signal is generated and sent out to activate a particular timed operation identified by an event ID (“EID”) associated with the timing value in the compare register. The current timing value in the compare register is then discarded, and the next timing value in sequence is retrieved into the compare register. In a first embodiment, timing values are stored in a hardware stack.

Abstract: A switchless Automatic Call Distribution (“ACD”) system distributing incoming calls to call agents networked via a low-cost data network such as an ethernet. Standard POTS calls are received from the Public Switched Telephone Network (“PSTN”), whereupon a POTS/packet gateway digitizes the signal (if necessary) and compresses it, advantageously from &mgr;-law format (64 Kbps) down to approximately 5-6 Kbps. The POTS/packet gateway then converts the signal into a packetized format. Responsive to a call distribution algorithm identifying the next available agent by IP address, the caller's packetized voice signals are distributed to the agent over the ethernet. At the same time, Web-enabled database tools generate “documents,” advantageously in html, containing database information regarding the caller. These documents are then distributed to the agent over the ethernet.

Abstract: A transaction model compartmentalizing generic ("user device-independent") information into transactions between a system and a user. Navigation of a series of these transactions enables the system and the user to exchange information in an organized and logical fashion. Each transaction is represented by a user device-independent transaction template which, when executed, enables a predefined exchange of information between the system and the user in user device-independent form. System functions, such as decision branch control, calculation steps and/or data retrieval are interlaced as required between transaction templates and process information so as to further facilitate the information exchange between the system and user in a user device-independent environment. User device templates are then associated with each transaction template. By scripting and conditioning the device-independent information exchange in the transaction template to take place in various device-dependent forms (i.e.

Abstract: A system and method for establishing a data connection between a computer and a live agent selected from an agent pool. A computer user uses the computer to connect via a network with an information server. While browsing information on the server, the customer can request to speak with an unspecified live agent. An agent queuing manager receives the customer's request and selects an appropriate live agent from a continuously updated pool of available agents. Then, the agent queuing manager establishes a data connection between the live agent and the customer. The data connection is preferably an audio connection, although a real time text "chat" mode is also available. The agent queuing manager maintains the agent pool by monitoring when agents log in and out of agent terminals, as well as when the agents are engaged with a computer user.

Abstract: A method for transferring multimedia information in a system including a primary site 101 and at least one satellite site 102 coupled to the primary site 101 by a transmission medium 103. Information defining each of a selected plurality of objects composing a multimedia presentation is compressed at primary site 101 using a compression/decompression protocol corresponding to a type of each of the objects. The compressed information is transferred to satellite site 102 from primary site 101 via medium 103. Scene information describing the multimedia presentation to be made from the objects is also transferred from primary site 101. The compressed information defining each of the objects is decompressed at satellite 102 using the corresponding compression/decompression protocol. The multimedia presentation composed of the selected objects is generated at satellite site 102 from the decompressed information as directed by the received scene information.