Abstract: Machine learning techniques are described for generating recommendations using decision trees. A decision tree is generated based on training data that comprises multiple training instances, each of which comprises a feature value for each of multiple features and a label of a target variable. The multiple features correspond to attributes of multiple content delivery campaigns. Later, feature values of a content delivery campaign are received. The decision tree is traversed using the feature values to generate output. Based on the output, one or more recommendations are identified and the one or more recommendations are presented on a computing device.

Abstract: The present invention provides a hydroxyacyl-coenzyme A dehydrogenase gene, an acyl-coenzyme A thiolase gene, genetically engineered strains and a use thereof. The hydroxyacyl-coenzyme A dehydrogenase gene encodes a protein (i) or (ii) as follows: (i) having an amino acid sequence according to SEQ ID NO 2; (ii) derived by substituting, deleting or inserting one or more amino acids in the amino acid sequence defined by (i) and having the same function as that of the protein of (i). The present invention constructs genetically engineered Mycobacterium strains lacking of a hydroxyacyl-coenzyme A dehydrogenase gene or an acyl-coenzyme A thiolase gene, which are used in the preparation of steroidal compounds, such as 1,4-BNA, 4-BNA, 9-OH-BNA, etc . . .

Abstract: A method for localizing for GPS denied environments is provided with a plurality of broadcasting transmitters and a software-defined radio (SDR). A pre-processing stage and a query phase is executed for localizing an unknown location. The results from the preprocessing stage are made available to the query stage. A simulated power spectrum is computed for a region of interest during the pre-processing phase via the plurality of broadcasting transmitters. A transmitter location, a radius of influence, and a power contour plot of each of the plurality of broadcasting transmitters are used for computing the simulated power spectrum. Next, a peak-finding process, a filtering process, and a localization process is executed during the query phase in order to identify the unknown location.

Abstract: Method and system are provided to determine a member profile associated with a reference to a person in a publication and may be used beneficially to identify members of a social network in news articles. The system may be configured to detect a publication that was shared by a member of an on-line social networking system, determine that the publication includes a name entity, and identify a member profile in the on-line social networking system that represents a member referenced by the name entity. The process of identifying a member profile in the on-line social networking system that represents a member referenced by the name entity may include building a list of candidate profiles from the member profiles of the on-line social networking system and selecting a matching profile from the candidate profiles, utilizing one or more disambiguation techniques.

Abstract: Profiling systems and methods of creating and using user interest profiles are described. In some example embodiments, the method includes: creating a topic set which includes topics which are organized in a hierarchical structure which includes a plurality of topic levels including an upper topic level and a lower topic level, each topic in the lower topic level being a subtopic of at least one of the topics in the upper topic level; monitoring interest in a plurality of documents for a user to identify one or more documents-of-interest to the user; and based on the monitored interest for the user, creating an interest profile for the user by determining a measure of topical interest for the user for at least one of the topics at the upper topic level and for a subtopic of that topic, the subtopic being at the lower topic level.

Abstract: Methods and comment association systems for associating one or more comments with one or more primary electronic documents are described. In one aspect, the method comprises: identifying, at a comment association system, one or more key terms from at least a portion of the one or more primary electronic documents; identifying, at the comment association system, one or more comments associated with the identified key terms; determining, at the comment association system, whether an identified comment is sufficiently related to the one or more primary electronic documents by calculating one or more relation score for that identified comment and comparing the relation score to one or more threshold; and if the identified comment is sufficiently related to the one or more primary electronic documents, then associating the identified comment with the one or more primary electronic documents at the comment association system.

Abstract: Method and system are provided to determine a member profile associated with a reference to a person in a publication and may be used beneficially to identify members of a social network in news articles. The system may be configured to detect a publication that was shared by a member of an on-line social networking system, determine that the publication includes a name entity, and identify a member profile in the on-line social networking system that represents a member referenced by the name entity. The process of identifying a member profile in the on-line social networking system that represents a member referenced by the name entity may include building a list of candidate profiles from the member profiles of the on-line social networking system and selecting a matching profile from the candidate profiles, utilizing one or more disambiguation techniques.

Abstract: A phrase identification system and method are provided. The method comprises: identifying one or more phrase candidates in the electronic document; selecting one of the phrase candidates; numerically representing features of the selected phrase candidates to obtain a numeric feature representation associated with that phrase candidate; and inputting the numeric feature representation into a machine learning classifier, the machine learning classifier being configured to determine, based on each numeric feature representation, whether the phrase candidate associated with that numeric feature representation is a phrase.

Abstract: Method and system are provided to determine a member profile associated with a reference to a person in a publication and may be used beneficially to identify members of a social network in news articles. The system may be configured to detect a publication that was shared by a member of an on-line social networking system, determine that the publication includes a name entity, and identify a member profile in the on-line social networking system that represents a member referenced by the name entity. The process of identifying a member profile in the on-line social networking system that represents a member referenced by the name entity may include building a list of candidate profiles from the member profiles of the on-line social networking system and selecting a matching profile from the candidate profiles, utilizing one or more disambiguation techniques.

Abstract: A method and system for cleaning an electronic document are provided. The method comprises: identifying at least one sentence in the electronic document; numerically representing features of the sentence to obtain a numeric feature representation associated with the sentence; inputting the numeric feature representation into a machine learning classifier, the machine learning classifier being configured to determine, based on each numeric feature representation, whether the sentence associated with that numeric feature representation is a bad sentence; and removing sentences determined to be bad sentences from the electronic document to create a cleaned document.

Abstract: Profiling systems and methods of creating and using user interest profiles are described. In some example embodiments, the method includes: creating a topic set which includes topics which are organized in a hierarchical structure which includes a plurality of topic levels including an upper topic level and a lower topic level, each topic in the lower topic level being a subtopic of at least one of the topics in the upper topic level; monitoring interest in a plurality of documents for a user to identify one or more documents-of-interest to the user; and based on the monitored interest for the user, creating an interest profile for the user by determining a measure of topical interest for the user for at least one of the topics at the upper topic level and for a subtopic of that topic, the subtopic being at the lower topic level.

Abstract: Methods and comment association systems for associating one or more comments with one or more primary electronic documents are described. In one aspect, the method comprises: identifying, at a comment association system, one or more key terms from at least a portion of the one or more primary electronic documents; identifying, at the comment association system, one or more comments associated with the identified key terms; determining, at the comment association system, whether an identified comment is sufficiently related to the one or more primary electronic documents by calculating one or more relation score for that identified comment and comparing the relation score to one or more threshold; and if the identified comment is sufficiently related to the one or more primary electronic documents, then associating the identified comment with the one or more primary electronic documents at the comment association system.

Abstract: This invention provides a device and method for language model switching and adaptation, wherein the device comprises a notification manager which notifies a language model switching section of the current status information or the request for the language model of an destination application when the status of the destination application is changed; a language model switching section which selects one or more language models to be switched from a language model set according to the received current status information or the request; a LMB engine decodes a user input using the one or more selected language models; and a language model adaptation section which receives the decoded result and modifies the one or more selected language models based on the decoded result. Therefore, the user input is more accurate even if the language model switching section performs different switches among different language models and the performance of the language models are improved by the language model adaptation section.

Abstract: A method and system for cleaning an electronic document are provided. The method comprises: identifying at least one sentence in the electronic document; numerically representing features of the sentence to obtain a numeric feature representation associated with the sentence; inputting the numeric feature representation into a machine learning classifier, the machine learning classifier being configured to determine, based on each numeric feature representation, whether the sentence associated with that numeric feature representation is a bad sentence; and removing sentences determined to be bad sentences from the electronic document to create a cleaned document.

Abstract: A phrase identification system and method are provided. The method comprises: identifying one or more phrase candidates in the electronic document; selecting one of the phrase candidates; numerically representing features of the selected phrase candidates to obtain a numeric feature representation associated with that phrase candidate; and inputting the numeric feature representation into a machine learning classifier, the machine learning classifier being configured to determine, based on each numeric feature representation, whether the phrase candidate associated with that numeric feature representation is a phrase.

Abstract: The present invention provides an apparatus and a method for categorizing entities based on time-series relation graphs. In each of the time-series relation graphs within a prescribed time period, nodes represent entities, and links between the nodes represent entity relations in a corresponding time unit. The inventive apparatus for categorizing entities based on time-series relation graphs comprises: a time-series relation graph categorizing means for categorizing the nodes in each of the time-series relation graphs to generate a node category result for the corresponding time unit in time sequence; and a category result post-processing means for post-processing all the node category results for the corresponding time units in time sequence generated by the time-series relation graph categorizing means to generate finally categorized nodes.

Abstract: The present invention provides a relation mining apparatus and method for mining data for time-series relations and events among texts in various forms such as news, blogs, industrial reports and technical papers which may refer to various relations. According to the present invention, it is possible to automatically extract entity relation instances from a large amount of the texts as described above originating from the Internet or other mediums, mine for time-series entity relations, relation scores and entity importances in various categories based on the extracted instances, and finally extract important events therefrom. Also, according to the present invention, it is possible to perform calculating on the above extracted time-series relations for the corporation entities and business relations, so as to achieve an analysis on Five Forces. Further, it is also possible to present the result to final users by a visualizing module.

Abstract: This invention provides a device and method for language model switching and adaptation, wherein the device comprises a notification manager which notifies a language model switching section of the current status information or the request for the language model of an destination application when the status of the destination application is changed; a language model switching section which selects one or more language models to be switched from a language model set according to the received current status information or the request; a LMB engine decodes a user input using the one or more selected language models; and a language model adaptation section which receives the decoded result and modifies the one or more selected language models based on the decoded result. Therefore, the user input is more accurate even if the language model switching section performs different switches among different language models and the performance of the language models are improved by the language model adaptation section.

Abstract: This invention provides a dictionary learning method, said method comprising the steps of: learning a lexicon and a Statistical Language Model from an untagged corpus; integrating the lexicon, the Statistical Language Mode and subsidiary word encoding information into a small size dictionary. And this invention also provides an input method on a user terminal device using the dictionary with Part-of-Speech information and a Part-of-Speech Bi-gram Model added, and a user terminal device using the same. Therefore, sentence level prediction and word level prediction can be given by the user terminal device and the input is speeded up by using the dictionary which is searched by a Patricia Tree index of a dictionary index.