Abstract: A method, an apparatus, a computer device and a readable medium for knowledge hierarchical extraction of a text are disclosed. The method comprises: performing word segmentation on a designated text to obtain a word list, the word list including at least one word arranged in a sequence in the designated text; analyzing part-of-speech of each word in the word list in the designated text, to obtain a part-of-speech list corresponding to the word list; predicting a SPO triple included in the designated text according to the word list, the part-of-speech list and a pre-trained knowledge hierarchical extraction model. By the technical solutions, the SPO triple included in any designated text however loose its organization and structure is may be accurately extracted based on the pre-trained knowledge hierarchical extraction model. Compared to the prior art, the efficiency and accuracy of knowledge hierarchical extraction may be effectively improved.

Abstract: The present invention relates to a skin topical composition having an antipruritic efficacy, comprising a combination of (A) birch sap and (B) betaine, and the skin topical composition does not comprise water added as a separate component. The skin topical composition is a pharmaceutical composition or a cosmetic composition.

Abstract: Methods, systems, and devices for wireless communications are described that provide improved repetition techniques for autonomous uplink transmissions. Configuration information may be provided from a base station to a user equipment (UE) for autonomous uplink transmissions, that may indicate initial resources for an initial uplink transmission slot and retransmission resources for a number of retransmissions. Time resources within the retransmission resources may be configured to provide enhanced reliability, such as by being configured in non-contiguous slots to avoid one or more other transmission channels. Time resources within slots may also be configured to prevent persistent collisions with autonomous uplink transmissions or retransmissions of two or more different UEs.

Abstract: An aerosol-generating element for an aerosol-generating article or system is provided, the element including: a solid continuous matrix structure; and an aerosol-generating formulation dispersed within the structure, in which the formulation is trapped within the structure and releasable from the structure upon heating of the element, in which the structure is an alginate matrix, in which the formulation consists of: nicotine from 1% by weight to 4% by weight of a total weight of the element, glycerin from 60% by weight to 80% by weight of the total weight of the element, a carboxylic acid from 0.5% by weight to 4% by weight of the total weight of the element, and balance water, in which the formulation dispersed within the structure is at least about 80% by weight of a total weight of the element, and in which the element includes less than 15% by weight of water.

Abstract: A method of manufacturing a bulk acoustic wave filter is provided, including: forming an acoustic reflection air cavity, a sacrificial layer, a seed layer, a lower electrode layer and a piezoelectric layer of n resonators on a substrate in sequence, wherein n is greater than or equal to 2; taking N from 1 to n for respectively repeating following steps: forming an N-th metal hard mask layer, defining an effective area of a first resonator to an N-th resonator by using a photolithography process, removing the N-th metal hard mask layer outside the effective area of the first resonator to the N-th resonator, oxidizing the piezoelectric layer outside the effective area of the first resonator to the N-th resonator to form an N-th oxidized part of the piezoelectric layer, and etching the N-th oxidized part of the piezoelectric layer; removing the metal hard mask layer of the effective area of the first resonator to the N-th resonator, so as to form the piezoelectric layer having different thicknesses of the first

Abstract: Methods, systems, and devices for wireless communications are described. A user equipment (UE) may be configured to estimate, during a first time interval, at least one of a first interference level for a first radio access technology (RAT) caused at least in part by a second RAT, a second interference level for the second RAT caused at least in part by the first RAT, or a combination thereof. The UE may transmit a control message to a base station using the first RAT based on at least one of the first estimated interference level, the second estimated interference level, or a combination thereof. The UE may then communicate, based on the control message and during a second time interval, with the base station using the first RAT and with a wireless device using the second RAT, the second time interval being after the first time interval.

Abstract: An aerosol-generating element for an aerosol-generating article or system is provided, the aerosol-generating element including: a solid continuous matrix structure; and an aerosol-generating formulation dispersed within the structure, in which the formulation is trapped within the structure and releasable from the structure upon heating of the aerosol-generating element, in which the structure is a polymer matrix including one or more matrix-forming polymers, in which the formulation dispersed within the structure includes at least one alkaloid or cannabinoid compound and a polyhydric alcohol, in which a polyhydric alcohol content in the formulation dispersed within the structure accounts for at least 30 percent by weight based on a total weight of the aerosol-generating element, and in which the aerosol-generating element has an equivalent diameter of at least about 0.5 millimetres and an ovality from about 2 percent to about 30 percent.

Abstract: Some aspects described herein relate to considering data rates of multiple interfaces in determining whether to decode communications received over the interfaces and/or to considering the multiple interfaces in determining parameters for granting resources for one or more of the interfaces.

Abstract: A method and an apparatus for generating a model are provided. The method includes: acquiring a sample set including sample sentences and labeling knowledge corresponding thereto; and selecting a sample from the sample set, and performing following training steps: inputting a sample sentence into a first initial model to generate first prediction knowledge corresponding to the sample sentence; inputting the first prediction knowledge into a second initial model to generate a first prediction sentence corresponding to the first prediction knowledge; inputting labeling knowledge into the second initial model to generate a second prediction sentence corresponding to the labeling knowledge; inputting the second prediction sentence into the first initial model to generate a second prediction knowledge corresponding to the second prediction sentence; determining a first reward signal; and training, using a reinforcement learning method based on the first reward signal to obtain a first model.

Abstract: The present application discloses a method and a device for indicating a path loss reference signal, a terminal, a base station and a storage medium. The method for indicating a path loss reference signal includes: receiving an uplink transmission control signaling transmitted by a base station; determining a spatial domain relationship parameter of uplink transmission according to the uplink transmission control signaling; and determining a path loss reference signal used for uplink transmission power control according to the spatial domain relationship parameter of uplink transmission. According to the technical solutions provided by the embodiments, the path loss reference signal is determined according to the spatial domain relationship parameter, which saves signaling overhead, and improves configuration flexibility of a path loss reference signal used for uplink transmission power control.

Abstract: Described herein are embodiments for systems and methods to incorporate skip-gram convolution to extract non-consecutive local n-gram patterns for comprehensive information for varying text expressions. In one or more embodiments, one or more recurrent neural networks are employed to extract long-range features from localized level to sequential and global level via a chain-like architecture. Comprehensive experiments on large-scale datasets widely used for the text classification task were conducted to demonstrate the effectiveness of the presented deep skip-gram network embodiments. Performance evaluation on various datasets demonstrates that embodiments of the skip-gram network are powerful for general text classification task set. The skip-gram models are robust and may be generalized well on different datasets, even without tuning the hyper-parameters for specific dataset.

Abstract: An amplifier circuit structure can include an amplifier located in a main path, and a first switch located in a bypass. One end of a second switch is a signal output end of the amplifier circuit structure, and the other end of the second switch is configured to selectively connect to a signal output end of the bypass or a signal output end of the main path. The first and second switches are configured to control their respective operating states when a first instruction is received, such that the main path is connected to the signal input end and the signal output end of the amplifier circuit structure; and to control their respective operating states when a second instruction is received, such that the bypass is connected to the signal input end of the amplifier circuit structure and the signal output end of the amplifier circuit structure.

Abstract: A suspension for a skateboard, comprising a pair of steering knuckles on opposite sides of a chassis frame of the skateboard, a pair of upper control arms connected between the pair of steering knuckles and the chassis frame, a pair of lower control arms connected between the pair of steering knuckles and the chassis frame, and a first spring damper connected between the pair of upper control arms.

Abstract: Described herein are system and method embodiments to improve word representation learning. Embodiments of a probabilistic prior may seamlessly integrate statistical disentanglement with word embedding. Different from previous deterministic methods, word embedding may be taken as a probabilistic generative model, and it enables imposing a prior that may identify independent factors generating word representation vectors. The probabilistic prior not only enhances the representation of word embedding, but also improves the model's robustness and stability. Furthermore, embodiments of the disclosed method may be flexibly plugged in various word embedding models. Extensive experimental results show that embodiments of the presented method may improve word representation on different tasks.

Abstract: Embodiments of three-dimensional (3D) memory devices with a 3D NAND memory array having a plurality of pages, an on-die cache coupled to the memory array on a same chip and configured to cache a plurality of batches of program data between a host and the memory array, the on-die cache having SRAM cells, and a controller coupled to the on-die cache on the same chip. The controller is configured to check a status of an (N?2)th batch of program data, N being an integer equal to or greater than 2, program an (N?1)th batch of program data into respective pages in the 3D NAND memory array, and cache an Nth batch of program data in respective space in the on-die cache as a backup copy of the Nth batch of program data.

Abstract: Various aspects of the present disclosure generally relate to wireless communication. In some aspects, a user equipment (UE) may determine that signaling associated with a resource allocation is associated with a service type. The UE may identify a time-domain resource, for a transmission associated with the service type, in connection with determining that the signaling associated with the allocation is associated with the service type. Numerous other aspects are provided.

Abstract: Certain aspects of the present disclosure provide techniques for wireless communication. An exemplary method generally includes receiving an indication of a modulation and coding scheme (MCS) to use for transmitting information, wherein the indication of the MCS indicates an MCS index value corresponding to an entry in an MCS lookup table, determining, based on the MCS index value, a transport block size (TBS) to use for transmitting the information, wherein determining the TBS comprises receiving an explicit indication of the TBS to use for transmitting the information, and transmitting the information using the MCS and TBS.

Abstract: A POI valuation method, apparatus, device and computer storage medium are disclosed. The method comprises: obtaining information of first POIs with known values and information of second POIs with unknown values within a regional range; creating a valuation model which is configured to revaluate a first POI using values of surrounding POIs of the first POI, the surrounding POIs including other first POIs and second POIs within a predetermined range of distance from the first POI, and adjusting values of second POIs in the surrounding POIs using an error between a revaluated value of first POI and the known value of the first POI; training the valuation model until the error is minimized; obtaining the values of the second POIs from the valuation model. The solutions may reduce the requirement for manpower and improve the valuation efficiency as compared with manually valuation of POIs one by one.

Abstract: Existing research on cross-lingual retrieval cannot take good advantage of large-scale pretrained language models, such as multilingual BERT and XLM. The absence of cross-lingual passage-level relevance data for finetuning and the lack of query-document style pretraining are some of the key factors of this issue. Accordingly, embodiments of two novel retrieval-oriented pretraining tasks are presented herein to further pretrain cross-lingual language models for downstream retrieval tasks, such as cross-lingual ad-hoc retrieval (CUR) and cross-lingual question answering (CLQA). In one or more embodiments, distant supervision data was constructed from multilingual texts using section alignment to support retrieval-oriented language model pretraining. In one or more embodiments, directly finetuning language models on part of an evaluation collection was performed by making Transformers capable of accepting longer sequences.

Abstract: According to embodiments of the present disclosure, a method and an apparatus for responding to a question, and a storage medium are provided. The method includes: determining a question characteristic representation corresponding to a question for an object; determining a comment characteristic representation corresponding to a first comment for the object; generating a first target characteristic representation by utilizing the question characteristic representation and the comment characteristic representation; and determining an answer for the question based on the first target characteristic representation.