Abstract: The present invention relates to text analysis, and discloses a text representation method. Aspects include identifying concepts in the text by using a knowledge base and determining relationship between the concepts and generating a concept graph by using the relationship between the concepts. Aspects also include determining connected components of the concept graph; calculating weight of the connected components and determining the concepts representing the text according to the weight of the connected components. By using correlation between concepts in a knowledge base and according to connected component theory of a graph, finds out a set of concepts which best represents subject of the text, and removes concepts irrelevant to the subject, thus improving accuracy of text representation and reducing noise.

Abstract: The present invention relates to text analysis, and discloses a text representation method. Aspects include identifying concepts in the text by using a knowledge base and determining relationship between the concepts and generating a concept graph by using the relationship between the concepts. Aspects also include determining connected components of the concept graph; calculating weight of the connected components and determining the concepts representing the text according to the weight of the connected components. By using correlation between concepts in a knowledge base and according to connected component theory of a graph, finds out a set of concepts which best represents subject of the text, and removes concepts irrelevant to the subject, thus improving accuracy of text representation and reducing noise.

Abstract: Disambiguation in mention detection. The method includes: determining at least one location in a text at which a target surface form in the text appears; obtaining an overall word-bag context of the target surface form in the text, the word-bag context at each of the at least one location including words within a predetermined neighborhood of the location; obtaining an overall resource context of the target surface form in the text, the resource context at each of the at least one location including resources corresponding to a further surface form within a predetermined neighborhood of the location; and determining a similarity between the target surface form and a candidate resource for the target surface form based on the overall word-bag context and the overall resource context. A system for disambiguation in mention detection is also provided.

Abstract: Methods and an apparatus for generating a clustering model and clustering based on the clustering model. The first method includes: extracting feature information of each of a plurality of historical messages in response to receiving the plurality of historical messages from a historical voice conversation; obtaining a correlation between the plurality of historical messages; and generating a clustering model that clusters the plurality of historical messages on the basis of the correlation and the feature information of each of the plurality of historical messages. A second method for using the clustering model is also provided. According to the present invention, a reliable and accurate clustering model is generated through which a plurality of messages is clustered and displayed on the basis of the clustering model.

Abstract: A document tagging method and apparatus. According to the method, a focused entity relevant to a basic document and a sentiment polarity of comments on the focused entity are acquired, and then a tag is generated on the basic document from the focused entity and the corresponding sentiment polarity. The present invention can tag the basic document with an opinion on the relevant focused entity and thus facilitate knowledge of the opinion on a relevant entity.

Abstract: A method and an apparatus for pushing specific content for a predetermined webpage, and a website server. The method for pushing specific content for text content on a predetermined webpage comprises: subjecting text content on a predetermined webpage to emotional analysis; determining a matching degree between a result of the emotional analysis and an emotion expressed by specific content to be pushed; and responding to that the matching degree determined above satisfies a predetermined condition, combining a part of the text content with the specific content to be pushed, thereby forming content to be pushed specific for users. By using the technology of the present invention, user can be avoided from feeling disgust for content to be pushed and accuracy of push can be enhanced.

Abstract: A method for generating a mapping rule for converting relational data into RDF format data includes obtaining a first mapping rule for converting relational data in a first relational database into RDF format data, wherein a URI pattern of the first mapping rule expresses features of the relational data in the first relational database; parsing a generated rule for the URI pattern of the first mapping rule; determining that a second relational model of the second relational database and a first relational model of the first relational database conform to pattern matching; and generating a second mapping rule for converting the relational data of the second relational database into RDF format data, based on a mapping relationship between the generated rule and the pattern matching of the first relational model and the second relational model.

Abstract: A method for generating a mapping rule for converting relational data into RDF format data includes obtaining a first mapping rule for converting relational data in a first relational database into RDF format data, wherein a URI pattern of the first mapping rule expresses features of the relational data in the first relational database; parsing a generated rule for the URI pattern of the first mapping rule; determining that a second relational model of the second relational database and a first relational model of the first relational database conform to pattern matching; and generating a second mapping rule for converting the relational data of the second relational database into RDF format data, based on a mapping relationship between the generated rule and the pattern matching of the first relational model and the second relational model.

Abstract: A method and apparatus for performing extended search are provided. The method includes receiving user-inputted keywords; extending the user-inputted keywords according to geographical information to acquire extended keywords; performing a search by using the extended keywords; and returning search results to the user. With the present technical solutions, privilege control can be effectively performed in a cloud storage system. With the present embodiments, more information may be provided to a user for reference.

Abstract: Method, apparatus, and hospital information system for prompting medication safety. The method includes: extracting identifier information and execution time of a prescribed drug from a treatment information database; according to the identifier information of the drug, extracting a record related to the drug from an information database; for each contributor factor of the prescribed drug, determining an unsafe time interval corresponding to the contributor factor of the drug; extracting identifier information and execution time of a target factor from the treatment information database; and, in response to the target factor matching with a contributor factor of the drug and at least a part of the execution time of the target factor falling within the unsafe time interval corresponding to the matched contributor factor, prompting medication safety in the complete drug taking lifetime. Through adopting the present invention, internal data process can be optimized and medication safety can be improved.

Abstract: A method and apparatus for performing extended search are provided. The method includes receiving user-inputted keywords; extending the user-inputted keywords according to geographical information to acquire extended keywords; performing a search by using the extended keywords; and returning search results to the user. With the present technical solutions, privilege control can be effectively performed in a cloud storage system. With the present embodiments, more information may be provided to a user for reference.

Abstract: A reference voltage generation circuit includes an auto-activation unit, an operational amplifier unit, and a tail current resistor. An input of the operational amplifier is grounded via the tail current resistor. The auto-activation unit is coupled to the operational amplifier so that the circuit operates at an operating point.

Abstract: Various embodiments of the invention describe nanoporous silicon (Si) network thin films with controllable porosity and thickness that are fabricated by a robust and scalable electrochemical process, and then released from Si wafers and transferred to flexible and conductive substrates. These nanoporous Si network thin films serve as high performance Li-ion battery electrodes, with an initial discharge capacity of 2570 mA h g?1, above 1000 mA h g?1 after 200 cycles without any electrolyte additives.

Abstract: A method and apparatus for performing extended search are provided. The method includes receiving user-inputted keywords; extending the user-inputted keywords according to geographical information to acquire extended keywords; performing a search by using the extended keywords; and returning search results to the user. With the present technical solutions, privilege control can be effectively performed in a cloud storage system. With the present embodiments, more information may be provided to a user for reference.

Abstract: A device for generating opinion pairs having sentiment orientation based impact relations, having: a reception unit which receives a plurality of object-oriented opinions; an extraction unit which extracts opinions having impact relations among the plurality of object-oriented opinions; an analysis unit which analyzes sentiment orientations of the extracted respective opinions; and a generation unit which establishes correlations between the opinions having sentiment orientation based impact relations to generate the opinion pairs having the sentiment orientation based impact relations. The present device can search and locate more implicit opinions as supplementary opinions by detecting impact relations from a plurality of opinions, and help users interested in different objects obtain more comprehensive comments and summaries.

Abstract: A method and apparatus for obtaining address update information. The method includes: obtaining webpage information related to an address update of a named entity; extracting keywords representing an address update clue of the named entity from the webpage information; determining the time and confidence of the address update clue; determining the address update information based on the time and confidence. Using the solution of the present invention can determine address update information through obtained website information.

Abstract: A method and apparatus for tracking a topic involved in a message in an interactive messaging environment containing a plurality of messages by combining with context messages associated with the message. The method includes the steps of: selecting a message from the plurality of messages contained in the interactive messaging environment; identifying context messages associated with the selected message from the plurality of messages contained in the interactive messaging environment; and grouping the selected message into one or more topic groups based on the identified context messages. The topic tracking method and apparatus of the invention can accurately determine similarities between messages in an interactive messaging environment such as microblog to perform topic tracking.

Abstract: A method for forming a solar cell having a plasmonic back reflector is disclosed. The method includes the formation of a nanoimprinted surface on which a metal electrode is conformally disposed. The surface structure of the nanoimprinted surface gives rise to a two-dimensional pattern of nanometer-scale features in the metal electrode enabling these features to collectively form the plasmonic back reflector.

Abstract: An optoelectronic device comprising an optically active layer that includes a plurality of domes is presented. The plurality of domes is arrayed in two dimensions having a periodicity in each dimension that is less than or comparable with the shortest wavelength in a spectral range of interest. By virtue of the plurality of domes, the optoelectronic device achieves high performance. A solar cell having high energy-conversion efficiency, improved absorption over the spectral range of interest, and an improved acceptance angle is presented as an exemplary device.

Abstract: A method and system for searching an electronic map. The method includes the steps of: receiving a query entered by a user for searching the electronic map; extracting place names in the query; extracting in web pages searched on a computer network co-occurrence place names that co-occurred with the place names; selecting extension place names from the co-occurrence place names, thereby obtaining a set of extension place names; generating a set of queries, wherein the set of queries includes the query entered by the user, and includes a query obtained by replacing the place names in the query entered by the user with each extension place name in the set of extension place names respectively; and sending the set of queries to a map search engine.