Abstract: Embodiments facilitate greater flexibility in definition of user segments for targeted advertising, by employing indexed semantic user profiles. Semantic user profiles are built through extraction of online user behavior data such as user search queries and page views, and include user interest information that is inferred based on user behavior. Semantic user profiles are then indexed to facilitate search for a set of users that fit specified semantic search terms. Search results for semantic profiles are ranked according to a ranking model developed through machine learning. In some embodiments, building and indexing of semantic profiles and learning of the ranking model is performed offline to facilitate more efficient online processing of queries.

Abstract: A knowledge extraction framework may iteratively enrich an ontology that is used to classify structured knowledge obtained from web pages based on structured knowledge previously acquired from other web pages. The framework may enable a user to define the ontology for extracting structured knowledge from a plurality of web pages. The framework applies the ontology using a supervised extraction algorithm to extract seed information from a set of web pages. The framework further applies an unsupervised extraction algorithm to extract the structured knowledge from an additional set of web pages. The framework subsequently maps the structured knowledge to the ontology based on the seed information to enrich the ontology.

Abstract: Embodiments facilitate greater flexibility in definition of user segments for targeted advertising, by employing indexed semantic user profiles. Semantic user profiles are built through extraction of online user behavior data such as user search queries and page views, and include user interest information that is inferred based on user behavior. Semantic user profiles are then indexed to facilitate search for a set of users that fit specified semantic search terms. Search results for semantic profiles are ranked according to a ranking model developed through machine learning. In some embodiments, building and indexing of semantic profiles and learning of the ranking model is performed offline to facilitate more efficient online processing of queries.

Abstract: Techniques are described for generating structured information from semi-structured web pages, and retrieving the structured knowledge in response to a user query that indicates a query intent. The structured information is automatically extracted offline from semi-structured web pages, through the use of an auto wrapper solution that is noise tolerant, scalable, and automatic. The structured information is stored in a knowledge base, and provided in response to a user search query that indicates a query intent. Extraction of structured information may also include clustering of pages based on their measured similarities. The clusters may be determined based on similar elements in the tag path text data of the pages. A minimum size threshold may be applied to the clusters.

Abstract: The related links recommendation technique described herein employs combined collaborative filtering to recommend related web pages to users. The technique creates multiple collaborative filters which are combined in order to create a combined collaborative filter to recommend web pages similar to a given web page to a user. One query-based collaborative filter is created by using query search clicks (e.g., user input device selection actions on search results returned in response to a search query). Another user-behavior-based collaborative filter is created by using query search clicks and user clicks while browsing websites (e.g., user input device selection actions while a user is browsing websites). Lastly, another content-based collaborative filter based on similar content of web pages is created by finding web pages with similar content.

Abstract: Embodiments facilitate greater flexibility in definition of user segments for targeted advertising, by employing indexed semantic user profiles. Semantic user profiles are built through extraction of online user behavior data such as user search queries and page views, and include user interest information that is inferred based on user behavior. Semantic user profiles are then indexed to facilitate search for a set of users that fit specified semantic search terms. Search results for semantic profiles are ranked according to a ranking model developed through machine learning. In some embodiments, building and indexing of semantic profiles and learning of the ranking model is performed offline to facilitate more efficient online processing of queries.

Abstract: One embodiment of the present invention provides a lock-and-key system for an electrical power plant. The system includes a plurality of locks and a smart key. A respective lock is installed with a standardized lock core, and the lock is associated with a lock identifier (ID). The smart key includes a key head that matches the standardized lock core; a lock-ID detector configured to detect the lock ID; a rotation stopper which, when enabled, is configured to prevent rotation of the key head while the key head is inserted into the standardized lock core; and a control module configured to disable the rotation stopper based on the detected lock ID, thereby facilitating the smart key to unlock the lock by rotating the key head.

Abstract: Architecture that defines domain knowledge on networks, such as the Internet, as tables where each row is an entity in the target domain and each column is an attribute of these entities. The corresponding values for entity-attribute pairs are the domain knowledge. The architecture provides semi-automatic and systematic ways to extract network knowledge from at least an unstructured and semi-structured network (the Internet), structuralizes the knowledge in table format, and uses the domain tables to build the online updated knowledge base.

Abstract: Various technologies pertaining to provision of graphical data to a client computing device responsive to receipt of a positional selection on a web page by a user of a client computing device are described herein. A computer-executable application executing on the client computing device detects that the user has selected a certain position on the web page, wherein this application is not called by code of the web page. The position is transmitted to an ad server, which conducts an auction for display space on the client computing device based at least in part upon the detection of the certain position.

Abstract: A classification process may reduce the computational resources and time required to collect and classify training data utilized to enable a user to effectively access online information. According to some implementations, training data is established by defining one or more seed queries and query patterns. A bi-partite graph may be constructed using the seed query and query pattern information. A traversal of the bi-partite graph can be performed to expand the training data to encompass sufficient data to perform classification of the present search task.

Abstract: Transfer of learning trains a new domain for the classification of search queries according to different tasks, as well as the generation of a corresponding domain-specific query classifier that may be used to classify the search queries according to the different tasks in the new domain. The transfer of learning may include preparing a new domain to receive classification knowledge from one or more source domains by populating the new domain with preliminary query patterns extracted for a search engine log. The transfer of learning may further include preparing the classification knowledge in each source domain for transfer to the new domain. The classification knowledge in each source domain may then be transferred to the new domain.

Abstract: A tool facilitating learning latent semantics for ranking (LLSR) tailored to the ranking task via leveraging relevance information of query-document pairs to learn a tailored latent semantic space such that other documents are better ranked for the queries in the subspace. The tool applying a learning latent semantics for ranking algorithm integrating LLSR, thereby enabling learning an optimal latent semantic space (LSS) for ranking by utilizing relevance information in the training process of subspace learning. The tool enabling an optimization of the LSS as a closed form solution and facilitating reporting the learned LSS.

Abstract: A method for capturing gaseous elemental mercury in a flue gas wherein a soluble mercury oxidizer is utilized to oxidize gaseous elemental mercury at a gas-liquid interface of a wet scrubber, and oxidized mercury product is removed prior to re-emission.

Abstract: The related links recommendation technique described herein employs combined collaborative filtering to recommend related web pages to users. The technique creates multiple collaborative filters which are combined in order to create a combined collaborative filter to recommend web pages similar to a given web page to a user. One query-based collaborative filter is created by using query search clicks (e.g., user input device selection actions on search results returned in response to a search query). Another user-behavior-based collaborative filter is created by using query search clicks and user clicks while browsing websites (e.g., user input device selection actions while a user is browsing websites). Lastly, another content-based collaborative filter based on similar content of web pages is created by finding web pages with similar content.

Abstract: A method and apparatus for utilizing an amine-based absorption compound for flue gas, or combustion gas, CO2 capture, according to embodiments of the present invention. In some embodiments, CO2 is delivered from a wet, or dry, flue gas scrubber to an absorber. The amine-based absorption compound is introduced in an aqueous solution to the absorber, which is located downstream of the wet, or dry, flue gas scrubber. In some embodiments, the CO2-laden amine-based absorption compound is then subjected to a regeneration step to regenerate the CO2-laden amine-based absorption compound and strip out absorbed CO2. In some embodiments, energy is supplied to a stripper through a reboiler to accomplish the regeneration step.

Abstract: A method for merging really simple syndication (RSS) feeds. Stories containing one or more terms may be merged into one or more clusters based on one or more links between the stories. A cluster frequency with which the terms occur in each cluster may be determined. A diameter for each cluster may be determined. A cluster that is most similar to one of the clusters may be determined based on the cluster frequency. The most similar cluster with the one of the clusters may be determined based on each diameter, and each cluster frequency.

Abstract: A method and system for use in information retrieval includes, for each of a plurality of terms, selecting a predetermined number of top scoring documents for the term to form a corresponding document set for the term. When a plurality of terms are received, optionally as a query, the system ranks, using an inverse document frequency algorithm, the plurality of terms for importance based on the document sets for the plurality of terms. Then a number of ranked terms are selected based on importance and a union set is formed based on the document sets associated with the selected number of ranked terms.

Abstract: A method for categorizing online user behavior data, including creating a target set of users based on an advertiser query, identifying two or more users in the target set having one or more first similar behavior attributes using a Minhash algorithm; and modifying the target set according to the two or more identified users.

Abstract: A method for capturing gaseous elemental mercury in a flue gas wherein a soluble mercury oxidizer is utilized to oxidize gaseous elemental mercury at a gas-liquid interface of a wet scrubber, and oxidized mercury product is removed prior to re-emission.

Abstract: Described is an efficient retrieval mechanism that quickly locates documents (e.g., corresponding to online advertisements) based on query term discrimination. A topmost subset (e.g., two) of search terms is selected according to their ranked importance, e.g., as ranked by inverted document frequency. The topmost terms are then used to narrow the number of rows of an inverted query index that are searched to find document identifiers and associated scores, such as computed offline by a BM25 algorithm. For example, for each document identifier of each important term, a fast search within each of the narrowed subset of rows (that also contain that document identifier) may be performed by comparing document identifiers to jump a pointer within each other row, followed by a binary search to locate a particular document. The scores of the set of particular documents may then be used to rank their relative importance for returning as results.