Abstract: Architecture that scales up the non-negative matrix factorization (NMF) technique to a distributed NMF (denoted DNMF) to handle large matrices, for example, on a web scale that can include millions and billions of data points. To analyze web-scale data, DNMF is applied through parallelism on distributed computer clusters, for example, with thousands of machines. In order to maximize the parallelism and data locality, matrices are partitioned in the short dimension. The probabilistic DNMF can employ not only Gaussian and Poisson NMF techniques, but also exponential NMF for modeling web dyadic data (e.g., dwell time of a user on browsed web pages).

Abstract: An input data set is treated as a plurality of grouped sets of key/value pairs, which enhances the utility of the MapReduce programming methodology. By utilizing such a grouping, map processing can be carried out independently on two or more related but possibly heterogeneous datasets (e.g., related by being characterized by a common primary key). The intermediate results of the map processing (key/value pairs) for a particular key can be processed together in a single reduce function by applying a different iterator to intermediate values for each group. Different iterators can be arranged inside reduce functions in ways however desired.

Abstract: A light emitting device comprising a first semiconductor layer, a second semiconductor layer and a quantum well layer, wherein the first semiconductor layer and the second semiconductor layer are disposed on the opposite sides of the quantum well layer, the quantum well layer comprising a plurality of quantum well rods which are separated from each other, and each of the quantum well rods has only one quantum well.

Abstract: An illuminating brick includes a block and at least one light-emitting element mounted in the brick. The brick has a top face, a bottom face and a plurality of lateral side surfaces interconnecting the top and bottom faces. The at least one light-emitting element is engaged in and optically coupled to at least one of the bottom face and lateral side surfaces. The lateral side surfaces and the bottom face are configured for reflecting and directing light emitted from the at least a light-emitting element to exit through the top face.

Abstract: Architecture that scales up the non-negative matrix factorization (NMF) technique to a distributed NMF (denoted DNMF) to handle large matrices, for example, on a web scale that can include millions and billions of data points. To analyze web-scale data, DNMF is applied through parallelism on distributed computer clusters, for example, with thousands of machines. In order to maximize the parallelism and data locality, matrices are partitioned in the short dimension. The probabilistic DNMF can employ not only Gaussian and Poisson NMF techniques, but also exponential NMF for modeling web dyadic data (e.g., dwell time of a user on browsed web pages).

Abstract: Architecture that automatically detects historical search contexts as well as behaviors related to a search query. Machine learning and hand-authored rules are employed to automatically identify search contexts. Historical information likely to be useful in the current context is surfaced. When a user enters a search query or executes another search behavior, past behaviors are exposed which are contextually related to the current behavior. The architecture also provides automatic discovery of historical contexts, features related to the contexts, and training or authoring of a system for classifying behavior into contexts, using some combination of the machine learning and/or hand-authored rules. A runtime system classifies the current user behavior into a context and surfaces contextual information to the user.

Abstract: A planar illumination device includes at least one light source, a housing structure, at least one light guide plate, and at least one light exiting surface. The at least one light source is received in the housing structure. The housing structure includes at least one light emitting surface through which the light emitted from the at least one light source exits the at least one housing. The at least one light guide plate includes at least one light incidence surface through which the light enters into the at least one light guide plate, and contacts with the at least one light emitting surface. The light exits the at least one light guide plate through the at least one light exiting surface. The at least one light exiting surface includes a plurality of continuously connected bulge points formed thereon.

Abstract: A light emitting diode device is provided, which comprises a substrate comprising a first growth surface and a bottom surface opposite to the first growth surface; a dielectric layer with a plurality of openings therein formed on the first growth surface; a plurality of semiconductor nano-scaled structures formed on the substrate protruding through the openings; a layer formed on the plurality of semiconductor nano-scaled structures with a second growth surface substantially parallel with the bottom surface; a light emitting diode structure formed on the second growth surface; wherein the diameters of the openings are smaller than 250 nm, and wherein the diameters of the plurality semiconductor nano-scaled structures are larger than the diameters of the corresponding openings.

Abstract: A light emitting diode (LED) illumination device (10) includes an LED (11), a frustum-shaped light guide member (12) and a light diffusing plate (13). The frustum-shaped light guide member (12) has a light input surface (120) and a light output surface (122) opposite to the light input surface (120). The light guide member (12) tapers from the light output surface (122) to the light input surface (120). The light input surface (120) is optically coupled to the LED (11), and the light diffusing plate (13) is optically coupled to the light output surface (122) of the light guide member (12).

Abstract: The application illustrates a light-emitting device including a contact layer and a current spreading layer on the contact layer. A part of the contact layer is a rough structure and a part of the contact layer is a flat structure. A part of the current spreading layer is a rough structure and a part of the current spreading layer is a flat structure. The rough region of the contact layer and the rough region of the current spreading layer are substantially overlapped.

Abstract: A planar illumination device includes at least one light source, a housing structure, at least one light guide plate, and at least one light exiting surface. The at least one light source is received in the housing structure. The housing structure includes at least one light emitting surface through which the light emitted from the at least one light source exits the at least one housing. The at least one light guide plate includes at least one light incidence surface through which the light enters into the at least one light guide plate, and contacts with the at least one light emitting surface. The light exits the at least one light guide plate through the at least one light exiting surface. The at least one light exiting surface includes a plurality of continuously connected bulge points formed thereon.

Abstract: An illumination device includes a light emitting module (20) and a mounting housing (10) for receiving the light emitting module therein. The light emitting module includes a light guide plate (21) and a light source (23) optically coupled to the light guide plate. The mounting housing includes a top plate (12), a mounting plate (14) and a side plate (16) interconnected between the top plate and the mounting plate. A receiving space (15) is cooperatively formed by the top plate, the mounting plate and the side plate of the mounting housing for receiving the light guide plate therein. A gap (120) is defined in the top plate above the light emitting module. An opening (160) is defined in one end of the side plate for allowing the light emitting module to insert into the receiving space.

Abstract: A method of processing relationships of at least two datasets is provided. For each of the datasets, a map-reduce subsystem is provided such that the data of that dataset is mapped to corresponding intermediate data for that dataset. The intermediate data for that dataset is reduced to a set of reduced intermediate data for that dataset. Data corresponding to the sets of reduced intermediate data are merged, in accordance with a merge condition. In some examples, data being merged may include the output of one or more other mergers. That is, generally, merge functions may be flexibly placed among various map-reduce subsystems and, as such, the basic map-reduce architecture may be advantageously modified to process multiple relational datasets using, for example, clusters of computing devices.

Abstract: A light emitting diode (LED) illumination device (10) includes an LED (11), a frustum-shaped light guide member (12) and a light diffusing plate (13). The frustum-shaped light guide member (12) has a light input surface (120) and a light output surface (122) opposite to the light input surface (120). The light guide member (12) tapers from the light output surface (122) to the light input surface (120). The light input surface (120) is optically coupled to the LED (11), and the light diffusing plate (13) is optically coupled to the light output surface (122) of the light guide member (12).

Abstract: A planar illumination device includes at least one light source, a housing structure, at least one light guide plate, and at least one light exiting surface. The at least one light source is received in the housing structure. The housing structure includes at least one light emitting surface through which the light emitted from the at least one light source exits the at least one housing. The at least one light guide plate includes at least one light incidence surface through which the light enters into the at least one light guide plate, and contacts with the at least one light emitting surface. The light exits the at least one light guide plate through the at least one light exiting surface. The at least one light exiting surface includes a plurality of continuously connected bulge points formed thereon.

Abstract: An illuminating brick includes a block and at least one light-emitting element mounted in the brick. The brick has a top face, a bottom face and a plurality of lateral side surfaces interconnecting the top and bottom faces. The at least one light-emitting element is engaged in and optically coupled to at least one of the bottom face and lateral side surfaces. The lateral side surfaces and the bottom face are configured for reflecting and directing light emitted from the at least a light-emitting element to exit through the top face.

Abstract: A MapReduce architecture may be utilized for sequence alignment algorithm processing (such as BLAST or BLAST-like algorithms). In addition, a MapReduce architecture may be extended such that memory of the computing devices of a MapReduce-configured system may be shared between different jobs of sequence alignment and/or other bioinformatics algorithm processing, thereby reducing overhead associated with executing such jobs using the MapReduce-configured system.

Abstract: A method of processing relationships of at least two datasets is provided. For each of the datasets, a map-reduce subsystem is provided such that the data of that dataset is mapped to corresponding intermediate data for that dataset. The intermediate data for that dataset is reduced to a set of reduced intermediate data for that dataset. Data corresponding to the sets of reduced intermediate data are merged, in accordance with a merge condition. In some examples, data being merged may include the output of one or more other mergers. That is, generally, merge functions may be flexibly placed among various map-reduce subsystems and, as such, the basic map-reduce architecture may be advantageously modified to process multiple relational datasets using, for example, clusters of computing devices.

Abstract: An input data set is treated as a plurality of grouped sets of key/value pairs, which enhances the utility of the MapReduce programming methodology. By utilizing such a grouping, map processing can be carried out independently on two or more related but possibly heterogeneous datasets (e.g., related by being characterized by a common primary key). The intermediate results of the map processing (key/value pairs) for a particular key can be processed together in a single reduce function by applying a different iterator to intermediate values for each group. Different iterators can be arranged inside reduce functions in ways however desired.