Abstract: Provided is a method of generating data for estimating a three-dimensional (3D) pose of an object included in an input image, the method including acquiring the input image including at least one moving object, estimating location information about each joint of a plurality of joints of the object included in the input image using a prediction model pretrained to estimate the location information about each joint of the plurality of joints of the object, and generating animation data that represents a movement of the object based on the estimated location information.

Abstract: Provided is a method of generating data for estimating a three-dimensional (3D) pose of an object included in an input image, the method including acquiring the input image including at least one moving object, estimating location information about each joint of a plurality of joints of the object included in the input image using a prediction model pretrained to estimate the location information about each joint of the plurality of joints of the object, and generating animation data that represents a movement of the object based on the estimated location information.

Abstract: A method for routing calls suited to use in a call center includes receiving a call from a customer, extracting features from an utterance of the call, and, based on the extracted features, predicting a class and a complexity of a dialog to be conducted between the customer and an agent. With a routing model, a routing strategy is generated for steering the call to one of a plurality of types of agent (such as to a human or a virtual agent), based on the predicted class and complexity of the dialog and a cost assigned to the type of agent. A first of the plurality of types of agent is assigned a higher cost than a second of the types of agent. The routing strategy is output.

Abstract: A method for generating dialogs for learning a dialog policy includes, for each of at least one scenario, in which annotators in a pool of annotators serve as virtual agents and users, generating a respective dialog tree in which each path through the tree corresponds to a dialog and nodes of the tree correspond to dialog acts provided by the annotators. The generation includes computing a measure of uncertainty for nodes in the dialog tree, identifying a next node to be annotated, based on the measure of uncertainty, selecting an annotator from the pool to provide an annotation for the next node, receiving an annotation from the selected annotator for the next node, and generating a new node of the dialog tree based on the received annotation. A corpus of dialogs is generated from the dialog tree.

Abstract: A method for generating dialogs for learning a dialog policy includes, for each of at least one scenario, in which annotators in a pool of annotators serve as virtual agents and users, generating a respective dialog tree in which each path through the tree corresponds to a dialog and nodes of the tree correspond to dialog acts provided by the annotators. The generation includes computing a measure of uncertainty for nodes in the dialog tree, identifying a next node to be annotated, based on the measure of uncertainty, selecting an annotator from the pool to provide an annotation for the next node, receiving an annotation from the selected annotator for the next node, and generating a new node of the dialog tree based on the received annotation. A corpus of dialogs is generated from the dialog tree.

Abstract: A method for routing calls suited to use in a call center includes receiving a call from a customer, extracting features from an utterance of the call, and, based on the extracted features, predicting a class and a complexity of a dialogue to be conducted between the customer and an agent. With a routing model, a routing strategy is generated for steering the call to one of a plurality of types of agent (such as to a human or a virtual agent), based on the predicted class and complexity of the dialogue and a cost assigned to the type of agent. A first of the plurality of types of agent is assigned a higher cost than a second of the types of agent. The routing strategy is output.

Abstract: A system and method provide for shared access to a database in a semi-trusted platform. In the method, for each of a set of users, provision is made for regenerating a respective user key, based on a respective predefined user input, such as a hashed password. One or more of the users is authorized to have access to an encrypted database. For each of these, the method includes encrypting a key for the encrypted database with the respective user's user key to generate an encrypted database key. During a user session, one of the authorized users is provided with access to the encrypted database by decrypting the database key from the encrypted database key with the respective user's user key, and decrypting the database, from the encrypted database, with the database key. The database key and each user's user key are not stored on the platform and are thus inaccessible to platform administrators and unauthorized users between user sessions.

Abstract: Methods and systems for dynamically training a web crawler. The web crawler maintains one or more categories each comprising a set of words. The method includes selecting at least one hyperlink in response to a query received from a user. The method further includes determining a hyperlink score for the at least one hyperlink based on a category score associated with each of one or more categories. The category score associated with each of the one or more categories is updated based at least in part on the hyperlink score. The updated category score is compared with the hyperlink score to select a category from the one or more categories. The set of words associated with the category is updated based on content of a web page pointed by the at least one hyperlink.

Abstract: A system and method for text language identification allow private information of a server and a client to be kept secret from each other. An encrypted score for each of a plurality of languages is received by the server from the client. The encrypted scores are generated by homomorphic addition of encrypted frequencies of n-grams in a list of n-grams extracted from text. The unencrypted list is not provided to the server. The encrypted frequencies of the n-grams in the list are extracted using encrypted resources which, for each of the plurality of languages, include an encrypted frequency for each of a set of n-grams. At the server, the encrypted scores are decrypted to generate unencrypted scores and information is provided to the client based on the unencrypted scores from which the client is able to identify a language for the text.

Abstract: A web page identified by a URL stored in a downloads queue is downloaded, and hyperlinks in the downloaded web page are identified. Each hyperlink is screened by parsing the hyperlink (optionally only the URL of the hyperlink) to identify features comprising character strings, computing for each feature values for one or more meta-features indicative of the hyperlinked web page being in a target language, aggregating the meta-feature values to generate a score for the hyperlink, and adding the URL of the hyperlink to the downloads queue conditional upon the score satisfying a screening criterion. The downloading, identifying, and screening are iteratively repeated to perform web crawling, and an index of web pages in the target language is constructed based on analysis of content of the downloaded web pages. The meta-features may include a transliterated target word meta-feature, a language code meta-feature, a country code meta-feature, or so forth.

Abstract: A system and method provide for shared access to a database in a semi-trusted platform. In the method, for each of a set of users, provision is made for regenerating a respective user key, based on a respective predefined user input, such as a hashed password. One or more of the users is authorized to have access to an encrypted database. For each of these, the method includes encrypting a key for the encrypted database with the respective user's user key to generate an encrypted database key. During a user session, one of the authorized users is provided with access to the encrypted database by decrypting the database key from the encrypted database key with the respective user's user key, and decrypting the database, from the encrypted database, with the database key. The database key and each user's user key are not stored on the platform and are thus inaccessible to platform administrators and unauthorized users between user sessions.

Abstract: Methods and systems for dynamically training a web crawler. The web crawler maintains one or more categories each comprising a set of words. The method includes selecting at least one hyperlink in response to a query received from a user. The method further includes determining a hyperlink score for the at least one hyperlink based on a category score associated with each of one or more categories. The category score associated with each of the one or more categories is updated based at least in part on the hyperlink score. The updated category score is compared with the hyperlink score to select a category from the one or more categories. The set of words associated with the category is updated based on content of a web page pointed by the at least one hyperlink.

Abstract: A web page identified by a URL stored in a downloads queue is downloaded, and hyperlinks in the downloaded web page are identified. Each hyperlink is screened by parsing the hyperlink (optionally only the URL of the hyperlink) to identify features comprising character strings, computing for each feature values for one or more meta-features indicative of the hyperlinked web page being in a target language, aggregating the meta-feature values to generate a score for the hyperlink, and adding the URL of the hyperlink to the downloads queue conditional upon the score satisfying a screening criterion. The downloading, identifying, and screening are iteratively repeated to perform web crawling, and an index of web pages in the target language is constructed based on analysis of content of the downloaded web pages. The meta-features may include a transliterated target word meta-feature, a language code meta-feature, a country code meta-feature, or so forth.

Abstract: A system and method for document management are provided. The method relies on a logging system which automatically generates image logs for input documents for each job (print, copy, fax, scan, etc.) processed by the multifunction printing device(s) of an organization. The image logs are processed to identify keywords which are the basis of a search for similar documents among those which have been previously archived as well as documents in other accessible document repositories, including Web documents. The method identifies matching documents and optionally also revisions and related documents. A procedure is provided for ensuring that for each document processed by a multifunction device or other image output device of the organization, image data is archived (or identified as a public document without archiving). The method avoids duplication by using a digital matching document, where available, enabling the images of the image log for the input document to be discarded.

Abstract: In a method of generating hardcopy output, a print job description is provided to a print operation. The print job is in a native format where the print job includes the print operation in the native format and the print job description. A modified print operation is formed in a meta-format in response to the print job description. The modified print operation is submitted in the native format to a print engine to perform the print operation.

Abstract: A system and method for document management are provided. The method relies on a logging system which automatically generates image logs for input documents for each job (print, copy, fax, scan, etc.) processed by the multifunction printing device(s) of an organization. The image logs are processed to identify keywords which are the basis of a search for similar documents among those which have been previously archived as well as documents in other accessible document repositories, including Web documents. The method identifies matching documents and optionally also revisions and related documents. A procedure is provided for ensuring that for each document processed by a multifunction device or other image output device of the organization, image data is archived (or identified as a public document without archiving). The method avoids duplication by using a digital matching document, where available, enabling the images of the image log for the input document to be discarded.

Abstract: In accordance with the disclosure, there is provided a method for identifying duplicate documents comprising drafting a first document and creating a near unique representative string based on the document content. The method further comprises searching for other documents with the same NRS and selectively assigning a duplicate group identification to the first document, the duplicate group identification is unique if no near unique representative string matches are found, or the duplicate group identification is the same as an associated duplicate document's duplicate group identification that matches the NRS. The method further comprises placing the DGI into a meta-data of the first document and recalling a list of duplicates of a particular document based upon user demand by searching the meta-data and selecting documents using the same DGI.

Abstract: In accordance with the disclosure, there is provided a method for identifying duplicate documents comprising drafting a first document and creating a near unique representative string based on the document content. The method further comprises searching for other documents with the same NRS and selectively assigning a duplicate group identification to the first document, the duplicate group identification is unique if no near unique representative string matches are found, or the duplicate group identification is the same as an associated duplicate document's duplicate group identification that matches the NRS. The method further comprises placing the DGI into a meta-data of the first document and recalling a list of duplicates of a particular document based upon user demand by searching the meta-data and selecting documents using the same DGI.

Abstract: In a method of generating hardcopy output, a printjob description is provided to a print operation. The print job is in a native format where the print job includes the print operation in the native format and the print job description. A modified print operation is formed in a meta-format in response to the print job description. The modified print operation is submitted in the native format to a print engine to perform the print operation.