Xerox Corporation

Consona Crm Inc.

Kanisa, Inc.

Google Inc.

Nuance Communications, Inc.

Knova Software Inc.

Max Copperman

Scott B Huffman

Marc Dymetman

David Kay

Mark A Angel

Raya Fratkina

Jeffrey H Rudy

Monica Anderson

Robert Stern

Gregory T Grefenstette

Francois Pacull

Susan Dawnstarr Daniel

Tiago Goncalves Cabaco

Joshua Andrews

Katherine Vadella

Tanya Kraljic

Gabrielle Gayles

Shanjian Li

Jieyu Lu

Luchuan Xu

Andreas Bode

Igor Ukrainczyk

Implementations relate to effectively localizing system responses, that include dynamic information, to target language(s), such that the system responses are grammatical and/or natural in the target language(s). Some of those implementations relate to various techniques for resource efficient generation of templates for a target language. Some versions of those implementations relate to resource efficient generation of target language natural language generation (NLG) templates and, more particularly, to techniques that enable a human user to generate a target language NLG template more efficiently and/or with greater accuracy. The more efficient target language NLG template generation enables less utilization of various client device resources and/or can mitigate the risk of flawed NLG templates being provided for live use in one or more systems.

Interactive graphical interfaces for efficient localization of natural language generation responses, resulting in natural and grammatical target language output

In accordance with aspects of the disclosure, a computing device may provide a user interface for developing an interactive natural-language response system, which may include a virtual assistant. A user may interact with a system using spoken, written (e.g., text), or other input methods. The user interface may allow a user to associate sentences with intents, tag words within the sentences with concepts, and construct a grammar using the associated intents and tagged concepts. The system may use the grammar for automatically predictively associating sentences with intents and words with concepts. The system may display in the foam of a chat transcript a single branch of a tree of a discussion between the virtual assistant and a user. The user interface may graphically display variable values to assist a user to test system responses under different simulated conditions.

Virtual assistant development system

A method and system are disclosed for retrieving information through the use of a multi-stage interaction with a client to identify particular knowledge content associated with a knowledge map. The present invention is an application program running on a server accessed via the world-wide web or other data network using standard Internet protocols, a web browser and web server software. In addition to an automated portion, the present invention allows a human dialog designer to model the way the system elicits information, giving a human feel to the dialog and a better customer experience. In operation, users start a dialog by directing their web browser to a designated web page. This web page asks the user some initial questions that are then passed to a dialog engine. The dialog engine then applies its methods and algorithms to a knowledge map, using dialog control information\ and the user's responses to provide feedback to the user. The feedback may include follow-up questions, relevant documents, and instructions to the user (e.g., instructions to contact a human customer service representative). This dialog engine response is rendered as a web page and returned to the user's web browser. The user can then respond further to the follow-up questions he or she is presented, and the cycle repeats. The invention can be implemented so that it can interact with customers through a wide variety of communication channels including the Internet, wireless devices (e.g., telephone, pager, etc.), handheld devices such as a Personal Data Assistant (PDA), email, and via a telephone where the automated system is delivered using an interactive voice response (IVR) and/or speech-recognition system.

System and method for providing an intelligent multi-step dialog with a user

A method and system organize and retrieve information using taxonomies, a document classifier, and an autocontextualizer. Documents (or other knowledge containers) in an organization and retrieval subsystem may be manually or automatically classified into taxonomies. Documents are transformed from clear text into a structured record. Automatically constructed indexes help identify when the structured record is an appropriate response to a query. An automatic term extractor creates a list of terms indicative of the documents' subject matter. A subject matter expert identifies the terms relevant to the taxonomies. A term analysis system assigns the relevant terms to one or more taxonomies, and a suitable algorithm is then used to determine the relatedness between each list of terms and its associated taxonomy. The system then clusters documents for each taxonomy in accordance with the weights ascribed to the terms in the taxonomy's list and a directed acyclic graph (DAG) structure is created.

System and method for implementing a knowledge management system

A method and system are disclosed for retrieving information through the use of a multi-stage interaction with a client to identify particular knowledge content associated with a knowledge map.

This document discusses, among other things, systems and methods for searching for relevant documents in a document corpus. Using, among other things, text provided by a user's query, the system undertakes a dynamic search process that includes at least one ordered sequence of searches. The results of a first search are evaluated to determine how to formulate a second or subsequent search, whether to perform a second or subsequent search, or whether or how to present to the user results from the search or searches performed up to that point. In one example, the first search uses tight criteria in conjunction with the language of the user's query. If the number of documents in the search results do not meet or exceed a threshold value, the search criteria is progressively loosened over subsequent searches. The search list may depend on, among other things, a characteristic of the user query or upon a result returned by a previous search on the user query.

Text search ordered along one or more dimensions

A method is provided for automatically classifying text into categories. In operation, a plurality of tokens or features are manually or automatically associated with each category. A weight is then coupled to each feature, wherein the weight indicates a degree of association between the feature and the category. Next, a document is parsed into a plurality of unique tokens with associated counts, wherein the counts are indicative of the number of times the feature appears in the document. A category score representative of a sum of products of each feature count in the document times the corresponding feature weight in the category for each document is then computed. Next, the category scores are sorted by perspective, and a document is classified into a particular category, provided the category score exceeds a predetermined threshold.

System and method for automatically classifying text

Automatic actions can be obtained through a network using an area of marking medium with machine-readable markings that encode an action/medium identifier. The action/medium identifier identifies an action that can be produced through the network, and also identifies the area of marking medium. For example, it may include a globally unique or network-wide page identifier as well as an action identifier that can be used to produce an action described by data associated with a counterpart digital page. Or it can include both a page identifier and a location identifier, with the location identifier also identifying an action that relates to the page's digital counterpart. Or it can include a document identifier and an action identifier. Or it can be a globally unique or network-wide sticker identifier that can be used to identify a document, a peripheral device, or another object to which the sticker is attached, and that also produces an action through the network.

Marking medium area with encoded identifier for producing action through network

A method and system organize and retrieve information using taxonomies, a document classifier, and an autocontextualizer. Documents (or other knowledge containers) in an organization and retrieval subsystem may be manually or automatically classified into taxonomies. Documents are transformed from clear text into a structured record. Automatically constructed indexes help identify when the structured record is an appropriate response to a query. An automatic term extractor creates a list of terms indicative of the documents' subject matter. A subject matter expert identifies the terms relevant to the taxonomies. A term analysis system assigns the relevant terns to one or more taxonomies, and a suitable algorithm is then used to determine the relatedness between each list of terms and its associated taxonomy. The system then clusters documents for each taxonomy in accordance with the weights ascribed to the terms in the taxonomy's list and a directed acyclic graph (DAG) structure is created.

Annotating activities are associated with portions of recordings by activity/recording data such as a table in which annotation identifiers are paired with timestamps. To obtain activity/recording data, an image input device, such as a manually positionable scanner or a fixed position camera, can produce an image signal set with information about a manual annotating activity and a recording/playback device can provide a recording portion signal, such as a timestamp, with information about a portion of a recording. The image signal set and the recording portion signal can then be used to automatically obtain the activity/recording data, associating information about the manual annotating activity with information identifying the portion of the recording. The information about the manual annotating activity can be an annotation identifier that includes information about the position or shape of the annotation. The activity/recording data can also include, for each annotation identifier, an image of the annotation. Activity/recording data can be used to playback a portion of a recording. Another image signal set can be obtained with information about a manual request activity, for example, and if the manual request activity is sufficiently similar to a manual annotating activity, the associated portion can be played back. Or a user can provide a selection signal indicating one of the annotations to obtain playback of the associated portion of the recording. If the annotating activity was performed with a keyboard or other user input device, the annotations can be printed, and the user can use a hand-held scanner to select an annotation to obtaining playback.

Obtaining and using data associating annotating activities with portions of recordings

A system and remote control device that controls an appliance, and obtains a desired function of the appliance, by invoking a remote procedure call in a remote server which will thereby actually control the appliance. The remote control device comprises a sensor; sensor responsive means for capturing information provided by the sensor, wherein the captured information at least contains parameters representative of the identification of the appliance, user-profile parameters, parameters representative of the address of the remote server, and a function name indicating the function to be performed by the appliance; marshalling means for encoding the captured information and for packaging the captured information into data in a remote procedure call format; and a transmitter for establishing a communication protocol with the remote server in order to transmit the packaged data to the remote server so that the remote server may execute the remote procedure call.

Generic handheld remote control device

An identifier of an action that can be produced through a network, such as a page identifier or other substrate identifier with a counterpart digital page accessible through a network, is used to obtain a network address. The identifier is provided to a first machine such as a router, and the first machine uses the identifier to obtain a first network address that is for a set of identifier values that includes the identifier's value. The first network address is then used to provide the identifier to a second machine such as a server. The second machine uses the identifier to obtain a second network address that is for the value of the identifier. The second network address, when provided on the network, produces the action identified by the value of the identifier. The identifier can be read from a coded substrate by image input circuitry and the image signals can be used by a third machine, such as a processor in a pointer, to provide the identifier to the first machine. The pointer's processor can also cache network addresses obtained from the first machine for use with subsequent identifiers in the same set.

Growing community of inventors

Santa Cruz, CA, United States of America

Max Copperman