Abstract: An impeller includes a hub and a plurality of blades. The blades are arranged around the hub, and each blade includes a leading edge, a blade tip, a root portion, a trailing edge, a windward side and a leeward side. The windward side including a first turning point and a second turning point, a first vertical height difference is formed from the blade tip to the first turning point, and a second vertical height difference is formed from the first turning point to the second turning point, and the first vertical height difference is greater than the second vertical height difference. The impeller apparently reduces the noise.

Abstract: The present disclosure describes a method of patterning a semiconductor wafer using extreme ultraviolet lithography (EUVL). The method includes receiving an EUVL mask that includes a substrate having a low temperature expansion material, a reflective multilayer over the substrate, a capping layer over the reflective multilayer, and an absorber layer over the capping layer. The method further includes patterning the absorber layer to form a trench on the EUVL mask, wherein the trench has a first width above a target width. The method further includes treating the EUVL mask with oxygen plasma to reduce the trench to a second width, wherein the second width is below the target width. The method may also include treating the EUVL mask with nitrogen plasma to protect the capping layer, wherein the treating of the EUVL mask with the nitrogen plasma expands the trench to a third width at the target width.

Abstract: A semiconductor structure is provided. The semiconductor structure includes a semiconductor wafer and a test structure. The semiconductor wafer has a substrate having a scribe line area, a first die area and a second die area. The first die area and the second die area are separated by the scribe line area extending along a first direction. The test structure is disposed in the scribe line area. The test structure includes a test device and a first test pad. The test device has a physical characteristic similar to a semiconductor device fabricated in the first die area or the second die area. The first test pad is electrically connected to the test device. A first distance between the first test pad and the first die area gradually increases from a center region to a peripheral region of the first test pad in the first direction.

Abstract: An electrical connector including: an insulative housing; four terminal modules received in the insulative housing and each having a row of terminals integrally formed with plural plastic fixing members, each terminal defining a contact portion, a tail portion opposite to the contact portion, and a holding portion therebetween; and a pair of metallic side plates fixing the four terminal modules together, each metallic side plates having plural holes, each plastic fixing member including a pair of positioning protrusions at two ends thereof to be received within corresponding holes, wherein the positioning protrusions respectively extend through the corresponding holes and are heat riveted to the metallic side plates.

Abstract: An electrical connector includes: an insulating housing having a front housing having a front top wall and a pair of front side walls and a rear housing cooperating with the front housing; and plural terminal modules accommodated in the insulating housing, each of the terminal modules including an insulating body and plural conductive terminals held in the insulating body, wherein the rear housing includes a rear top wall and a pair of rear side walls extending downward from the rear top wall, and the rear top wall and the pair of rear side walls surround the terminal module in three directions.

Abstract: A method for compressing a machine learning model by an electronic device. The method may comprise determining a compression parameter of a set hidden layer in a model based on a pruning number of respective channels included in the set hidden layer and a pruning loss of each hidden layer of the model; and compressing the model based on the compression parameter of the set hidden layer. The compression parameter may be related to a pruning of the model.

Abstract: A method, system, and device for selecting advertisement contents are provided. The method for selecting advertisement contents includes: receiving an advertisement request carrying advertisement request parameters; matching the advertisement request parameters with advertisement description information including a version field; and returning the matched advertisement description information, the matched version field in the advertisement description information, and an advertisement content corresponding to the matched advertisement description information. An advertisement includes multiple versions of advertisement description information, which enhances the flexibility in advertisement description and increases the opportunity of selecting and presenting the advertisement to users, thereby maximizing the benefit of the advertiser, saving the storage space, and reducing the amount of data transmission.

Abstract: The embodiments of the present invention disclose an advertisement measuring and reporting system and method. The system includes an advertisement client adapted to communicate with an advertisement server about advertisement; the advertisement client is adapted to obtain user interactions with advertisement, process the user interactions with advertisement according to an advertisement metrics policy, and report the processed user interactions with advertisement to the advertisement server; and the advertisement server is adapted to collect the processed user interactions with advertisement. Meanwhile, the embodiment of the present invention also provides a terminal, an advertisement client, an advertisement server, an advertisement measuring and reporting method and an advertisement interaction record collection method.

Abstract: A method for generating user attribute information includes: receiving event information which carries a time range, a space range and an event ID of an event; collecting users compliant with preset conditions according to the time range and space range in the event information to form a user collection; obtaining extended attributes of the event from the event information corresponding to the event ID; and generating attribute information of the users in the user collection correlated with the event according to the extended attributes of the event. An apparatus for generating user attribute information is also provided. The technical solution of the present invention reflects the characteristic attributes of the user accurately.