132 results about "Knowledge sources" patented technology

In language evaluation systems, user expressions are often evaluated by speech recognizers and language parsers, and among several possible translations, a highest-probability translation is selected and added to a dialogue sequence. However, such systems may exhibit inadequacies by discarding alternative translations that may initially exhibit a lower probability, but that may have a higher probability when evaluated in the full context of the dialogue, including subsequent expressions. Presented herein are techniques for communicating with a user by formulating a dialogue hypothesis set identifying hypothesis probabilities for a set of dialogue hypotheses, using generative and/or discriminative models, and repeatedly re-ranks the dialogue hypotheses based on subsequent expressions. Additionally, knowledge sources may inform a model-based with a pre-knowledge fetch that facilitates pruning of the hypothesis search space at an early stage, thereby enhancing the accuracy of language parsing while also reducing the latency of the expression evaluation and economizing computing resources.

A method of encoding knowledge is disclosed, which can be used to automatically detect problems in software application deployments. The method includes accessing a source of knowledge describing a problem known to occur in deployments of a particular software application, and which identifies a plurality of conditions associated with the problem. An encoded representation of the knowledge source is generated according to a predefined knowledge encoding methodology. The encoded representation is adapted to be applied automatically by a computer to analyze data representing a current state of a monitored deployment of the software application to detect whether the conditions and the problem exist therein. In various implementations, the encoded representation of the knowledge can include queries for deployment information, information concerning the relative importance of the conditions to a detection of the problem, and/or logical constructs for computing a confidence value in the existence of the problem and for determining whether to report the problem if some of the conditions are not true. The knowledge source can comprise a text document (such as a knowledge base article), a flowchart of a diagnostic troubleshooting method, and the like. Also disclosed are methods of at least partially automating the encoding process.

A method for fraud detection leverages an existing financial institution's fraud classification functionality, which produces a first level detection, with a “user-centric” classification functionality, which produces a “second” or more fine-grained detection regarding a potentially fraudulent transaction. After passing through an existing (“institution-centric”) fraud detection technique, a transaction that has been identified as potentially fraudulent is then subject to further analysis and classification at the “user” level, as it is the user is presumed to be the best source of knowledge of the legitimate credit card use. Information about the transaction is shared with the consumer, preferably via one or more near real-time mechanisms, such as SMS, email, or the like. Based on the user's response (or lack thereof, as the case may be), one or more business rules in the institution's fraud detection system can then take an appropriate action (e.g., no action, reverse the transaction if complete, deny the transaction if in-progress, or the like).

The subject invention relates to systems and methods that automatically combine or interleave received search results from across knowledge databases in a uniform and consistent manner. In one aspect, an automated search results blending system is provided. The system includes a search component that directs a query to at least two databases. A learning component is employed to rank or score search results that are received from the databases in response to the query. A blending component automatically interleaves or combines the results according to the rank in order to provide a consistent ranking system across differing knowledge sources and search tools.

The invention discloses an automatic question answering method and a system thereof. The automatic question answering method comprises the following steps: step A, the information of a question is acquired through the analysis of the question; step B, the query criterion is generated according to the analysis results of the question for retrieval by using the search engine in knowledge source, thus searching out the file collection relating to the query criterion; the file collection is divided into paragraph collections; step C, paragraph collections relating to the query criterion are retrieved in a centralized way from the paragraphs by using the paragraph retrieval method according to the file retrieval results; and the answer of the question is returned from the paragraph collections. The automatic question answering method and the system improve the accuracy of answers and the precision of returning correct results.

An “Iterative Query Reformulator” provides various techniques for using a computational engine to reformulate initial queries through one or more iterations. This query reformulation process ensures that results returned from search engines or recommendation systems using a reformulated query have improved relevance relative to results that would have been returned using only the initial query. More specifically, the Iterative Query Reformulator provides an end to end solution that uses computations from one or more knowledge databases or knowledge sources to find “partial answers” to subqueries derived or extracted from an initial query. These partial answers are then used to reformulate the initial query, with the reformulated query being used by the search engines or recommendations systems to provide results that are highly relevant to the initial query. Determinations of whether to continue reformulation iterations are based on evaluating user metrics from historical search logs having queries that match reformulated queries.

A method and program product for integrating different data sources has steps of obtaining semantic information from each of the different data sources (200, 202, 210), creating a conceptual model of (218, 220, 22) the data source using the semantic information, and accessing one or more secondary knowledge sources. The secondary information sources contain information regarding the relations of data from different of the databases, so that an integrated semantic model of all of the databases (200, 202, 210) may be created. Queries can then be processed using the integrated semantic model.

The invention discloses a method of recognizing voice. The method comprises steps: a preset voice knowledge source is used to generate searching space containing preset information of a client and used for decoding voice signals; a to-be-recognized voice signal feature vector sequence is extracted; the possibility that the feature vector is corresponding to a searching space basic unit is calculated; and with the possibility as input, decoding operation is executed in the searching space to obtain a word sequence corresponding to the feature vector sequence. The invention also provides a device for recognizing voice, and another method and another device for recognizing voice. By adopting the method provided by the invention, as the preset information of the client is included while the searching space for decoding is generated, information related to the client can be accurately recognized while the voice signals acquired by the client are recognized, and thus, the voice recognition accuracy can be improved, and the use experience of the user is enhanced.

A method, apparatus and computer program product for creating a dialog system for web content is described. Knowledge is extracted from a target web application for the dialog system. The knowledge includes an organizational structure of the target web application and domain knowledge pertinent to the target web application. A deep learning process associates the domain knowledge with the organization structure of the target application. A plurality of knowledge sources of different respective types are created from the domain knowledge and the organizational structure. Each of the knowledge sources is used for providing answers to user queries to the dialog system. As part of the invention, a semantic matcher is provided to select among the answers provided by the plurality of knowledge sources for a best answer to a user query.

A system and method are disclosed for obtaining a plurality of unlabeled text documents; obtaining an initial concept; obtaining keywords from a knowledge source based on the initial concept; scoring the plurality of unlabeled documents based at least in part on the initial keywords; determining a categorization of the documents based on the scores; performing a first feature selection and creating a first vector space representation of each document in a first category and a second category, the first and second categories based on the scores, the first vector space representation serving as one or more labels for an associated unlabeled textual document; and generating the training set including a subset of the obtained unlabeled textual documents, the subset of the obtained unlabeled documents including a documents belonging to the first category and documents belonging to the second category.

The invention provides a method and equipment for constructing a text classifier by referencing external knowledge. The method comprises the steps of: inputting a label text set; extracting internal characteristics of the label text set; constructing external characteristics of the label text set by referencing an external knowledge source (such as a dictionary); comprehensively considering the internal characteristics and the external characteristics of the label text set, and selecting training texts from the label text set; and learning the generation of the text classifier by using the selected training texts. According to the invention, sample distribution deviation generated by the label text set can be possibly regulated by the external characteristics automatically generated by the external knowledge source, and therefore, the finally trained classifier has better generalization capability and robustness.

In order to provide effective diagnostics and logging of error messages produced during the execution of processes across multiple components, systems and methods are disclosed for the generating, managing, and processing centralized logs containing those error messages. In particular, the components may write error messages to a centralized log instead of writing the error messages to local log files. These error messages may include exception messages and diagnostics messages. These various error messages in the centralized log can be read, identified, and organized. Furthermore, enrichments and/or analytics may be applied to the error messages based on information from a knowledge source or the application of one or more machine learning models. The organized error messages, enrichments, and analytics can be stored in an output log that can be easily retrieved and viewed through a graphical interface. The organized error messages, enrichments, and analytics may work together to allow for more effective diagnosing of execution errors.

The invention discloses a tax domain-oriented knowledge graph construction method. According to the construction method, a top-down and bottom-up combined mode is adopted. The method comprises the following steps: firstly, constructing a tax knowledge graph mode based on tax knowledge in a tax expert system; then, data processing including tax data source selection and acquisition, data cleaning and the like is carried out; then information extraction is carried out, and information extraction is carried out on the processed data according to different types and a mode diagram; tax knowledge fusion is carried out, knowledge sources in a tax knowledge graph are different, problems of knowledge repetition, relationship redundancy and the like need to be subjected to mode matching, entity alignment and the like, and knowledge is stored in a knowledge base after knowledge fusion is completed. And finally, knowledge feedback is carried out, and knowledge conflicts in intelligent tax model construction are solved by using a tax expert system. And finally, the problem of untight combination of the knowledge graph and the tax scene is solved.

The invention discloses a method for building a conceptual knowledge map based on Wikipedia, which includes steps: firstly, extracting knowledge documents from a Wikipedia full-text database and classifying and storing the extracted documents; secondly, analyzing each knowledge document to obtain knowledge information and storing the knowledge information in a database; thirdly, transforming the knowledge information into corresponding conceptual knowledge nodes and category knowledge nodes on the basis of standards in a knowledge node format description library and storing the knowledge nodes in file format locally; fourthly, building an index to provide a function of knowledge information searching via key words and displaying the searching results. The conceptual knowledge map is applicable to a plurality of fields, the knowledge information can be updated along with Wikipedia, obtaining knowledge source is easy, knowledge information is comprehensive, extensible markup language is adopted to describe the knowledge node information, and accordingly, the conceptual knowledge map is high in extensibility and can be applied to multiple platforms.

A method, apparatus and computer program product for presenting a user interface for a conversational system is described. A user input is received in a dialog between a user and the conversational system, the user input in a natural language. A domain trained semantic matcher is used to determine a set of entities and a user intent from the user input. One or more queries is generated to selected ones of a plurality of knowledge sources, the knowledge sources created from domain specific knowledge. The results from the one or more queries are ranked based on domain specific knowledge. A system response is presented in the dialog based on at least a highest ranked result from the plurality of knowledge sources.

Systems and methods are provided for improving language models for speech recognition by adapting knowledge sources utilized by the language models to session contexts. A knowledge source, such as a knowledge graph, is used to capture and model dynamic session context based on user interaction information from usage history, such as session logs, that is mapped to the knowledge source. From sequences of user interactions, higher level intent sequences may be determined and used to form models that anticipate similar intents but with different arguments including arguments that do not necessarily appear in the usage history. In this way, the session context models may be used to determine likely next interactions or "turns" from a user, given a previous turn or turns. Language models corresponding to the likely next turns are then interpolated and provided to improve recognition accuracy of the next turn received from the user.

The invention relates to a user interaction intention recognition method and system based on a speech act theory. The user interaction intention recognition method comprises the steps of constructing a behavior discourse marker dictionary based on an external knowledge source; automatically labeling the intention of an online text input by a user on a social media platform according to the behavior discourse marker dictionary; using an automatic tagged corpus to train a classifier based on features and conduct classified recognition on the intention of the online text to determine the interaction intention category of the user. According to the user interaction intention recognition method based on the speech act theory, by constructing the behavior discourse marker dictionary corresponding to different intention categories based on the external knowledge source, automatically labeling and expanding the corpus based on the behavior discourse marker dictionary, and conducting recognition on feature classification, the interaction intention of the user in a social media can be effectively recognized, the recognition accuracy is high, the user interaction intention recognition method and system can be used for intention analysis and recognition in the field of business intelligence, social situations, comment situations, decision evaluation and the like, and thus the application range is wide.

A method transforms unstructured text into structured data in a domain-specific ontology. The method includes recording an input block of text using an information extraction module (IEM), accessing a domain-specific ontology and supplemental data in a knowledge source(s) via the IEM, processing the input text block, and using the IEM to generate a plurality of nodes in the domain-specific ontology. Each node classifies the unstructured text to corresponding objects of interest, thereby transforming the unstructured text into the structured data. An IEM is also provided having a computer device and an algorithm executable thereby to transform unstructured text into structured data in a domain-specific ontology. The IEM is adapted for recording a text phrase using the computer device, accessing and retrieving the domain-specific ontology and supplemental data from a knowledge source(s), and processing the text block using the computer device to generate a plurality of nodes in the domain-specific ontology.