Abstract: A forecasting system is designed to train and utilize dynamic models for forecasting metrics associated with content provision. For example, a forecasting system may identify characteristics of received training data relating to a particular title. The characteristics may be used to select one of a plurality of available forecasting models to train.

Abstract: An artificial dressing, use of artificial dressing for promoting wound healing and method of manufacturing artificial dressing are disclosed. The artificial dressing comprises a gelatin, a polysorbate 20 and a glutaraldehyde for promoting wound healing. The manufacturing method of the artificial dressing comprises the steps of enabling the composition of the gelatin, the polysorbate 20 and the glutaraldehyde to perform a foaming step, a molding step and a freeing step in sequence. The artificial dressing has obvious pores and is capable of absorbing a liquid weight about 25-35 times of its body weight. After thermal disintegration experiments and adhesion tests, disintegration of the artificial dressing is not observed, and in animal experiments, compared with commercially available dressings, the artificial dressing of the invention is capable of accelerating wound healing when applied dryly; and capable of proliferating functional tissues of the wound when applied wetly.

Abstract: Provided are an anti-TSLP monoclonal antibody, an antigen-binding fragment thereof and a use thereof. The antibody has relatively high affinity with a TSLP antigen, may effectively inhibit the binding of the TSLP antigen to a receptor complex thereof, and then prevents a TSLP-targeted immune cell from releasing a pro-inflammatory cytokine, thereby asthma attack is prevented and asthma control is improved. The monoclonal antibody molecule obtained by screening in the present application also has relatively high thermal stability and relatively good safety. The present application may be used for treating asthma, chronic obstructive pulmonary disease, chronic eosinophilic pneumonia, idiopathic pulmonary fibrosis and allergic dermatitis; and the asthma includes severe asthma, eosinophilic or non-eosinophilic asthma, and low eosinophilic asthma.

Abstract: Systems, apparatuses, methods, and computer program products are disclosed for unsupervised named entity recognition. An example method includes receiving, by a communications circuitry, a reference named entity list, the reference named entity list identifying a set of named entities and an entity type of each identified named entity. The example method further includes generating, by a vectorizer, vectors from the named entities identified in the reference named entity list, and consolidating, by a synthesizer, the generated vectors into a set of representative vectors, wherein each representative vector is associated with a particular entity type. Finally, the example method receiving, by an analysis engine, a set of text, and performing, by the analysis engine, named entity recognition on the set of text using the set of representative vectors to generate a tagged set of text.

Abstract: In general, this disclosure describes techniques for generating and evaluating automatic transcripts of audio recordings containing human speech. In some examples, a computing system is configured to: generate transcripts of a plurality of audio recordings; determine an error rate for each transcript by comparing the transcript to a reference transcript of the audio recording; receive, for each transcript, a subjective ranking selected from a plurality of subjective rank categories; determine, based on the error rates and subjective rankings, objective rank categories defined by error-rate ranges; and assign an objective ranking to a new machine-generated transcript of a new audio recording, based on the objective rank categories and an error rate of the new machine-generated transcript.

Abstract: The present invention provides an artificial dressing and a use of the artificial dressing for promoting wound healing. The artificial dressing includes a gelatin and a fungal extract.

Abstract: In general, this disclosure describes techniques for generating and evaluating automatic transcripts of audio recordings containing human speech. In some examples, a computing system is configured to: generate transcripts of a plurality of audio recordings; determine an error rate for each transcript by comparing the transcript to a reference transcript of the audio recording; receive, for each transcript, a subjective ranking selected from a plurality of subjective rank categories; determine, based on the error rates and subjective rankings, objective rank categories defined by error-rate ranges; and assign an objective ranking to a new machine-generated transcript of a new audio recording, based on the objective rank categories and an error rate of the new machine-generated transcript.

Abstract: Provided are a fusion protein and a medical use thereof. The fusion protein is prepared by fusing a human interleukin-10 (IL-10) with an Fc fragment of an immunoglobulin IgG via a linker peptide. The Fc fragment is the Fc fragment of human IgGI, human IgG2 and human IgG4, and the human interleukin-10 includes an amino acid sequence as shown in SEQ ID NO: 1. The fusion protein may be used for the treatment of immunological diseases and cancers, and prolongs the in vivo half-life of human IL-10 and increases the in vivo stability of human IL-10.

Abstract: The present invention provides an artificial dressing and a use of the artificial dressing for promoting wound healing. The artificial dressing includes a gelatin and a fungal extract.

Abstract: The present application relates to the field of biomedicine, and particularly provides an anti-LAG-3 monoclonal antibody and an antigen binding fragment thereof, and use thereof, wherein the anti-LAG-3 monoclonal antibody and the antigen binding fragment thereof comprise a heavy chain variable region and a light chain variable region, and are selected from any one of A-1, A-2: A-3, and A-4.

Abstract: Disclosures of the present invention describe a double-layer dressing containing silk fibroin and a method for making the same, wherein the double-layer dressing mainly comprises a silk fibroin layer and a calcium-degradation silk fibroin layer connected to the silk fibroin layer, and it is worth emphasizing that, results of animal experiment have proved that this novel double-layer dressing is an outstanding hemostatic wound dressing; Moreover, additional adhesion, resulted from the solidification of tissue fluid, can be effectively prevented from forming between skin wound and wound dressing under the use of this double-layer dressing.

Abstract: Systems and methods for detecting demographic bias in automatic speech recognition (ASR) systems. Corpuses of transcriptions from different demographic groups are analyzed, where one of the groups is known to be susceptible to bias and another group is known not to be susceptible to bias. A difference between the transcription accuracy for the first group and a transcription accuracy for a second group is measured. ASR accuracy for each group is measured and compared to each other using both statistics-based and practicality-based methodologies to determine whether a given ASR system or model exhibits a meaningful level of bias. Based on the statistical significance and the practical significance, an alert including a recommendation to adjust the ASR model is generated.

Abstract: Systems and methods for detecting demographic bias in automatic speech recognition (ASR) systems. Corpuses of transcriptions from different demographic groups are analyzed, where one of the groups is known to be susceptible to bias and another group is known not to be susceptible to bias. ASR accuracy for each group is measured and compared to each other using both statistics-based and practicality-based methodologies to determine whether a given ASR system or model exhibits a meaningful level of bias.

Abstract: The present disclosure provides a method of measuring a plurality of voids in an underfill material of an underfill package. The method includes operations of obtaining a welding angle profile of the underfill package; obtaining a simulated void profile of the underfill package according to the welding angle profile; determining a plurality of high-risk void regions according to the simulated void profile; simulating, according to a selected pressure and a selected temperature of the underfill material, a first high-risk void region of the plurality of high-risk void regions to generate an updated void profile; and determining whether the updated void profile meets a void requirement of the underfill package.

Abstract: An anti-Epidermal Growth Factor Receptor (EGFR) and anti-Cluster-of-Differentiation (CD3) bispecific antibody and uses thereof are provided. The bispecific antibody can specifically bind to antigen epidermal growth factor receptor (EGFR) on tumor cell surface and antigen CD3 molecule on immune cell surface. The single-chain antibody (ScFv) of an anti-CD3 antibody is located at the C terminal of the constant region of the anti-EGFR antibody. Further provided are a preparation method of the bispecific antibody and clinical applications of the bispecific antibody. The bispecific antibody has high affinity, and is used for treating tumor diseases caused by high expression or abnormal expression of the EGFR and other diseases caused by other overexpression of the EGFR.

Abstract: The application discloses an anti-Proprotein Convertase Subtilisin/Kexin Type 9 (PCSK9) monoclonal antibody, an amino acid sequence comprising coded antibody variable regions and Complementarity-Determining Regions (CDR), the obtaining method, and application of monoclonal antibody. The anti-PCSK9 monoclonal antibodies are screened from a phage antibody library; affinity maturation is achieved by using a method that the phage antibody library is established by means of chain displacement; After screening the light chain CDR1, 2, 3 Mutant Library of monoclonal antibodies obtained from primary screening; monoclonal antibodies having high affinity are selected; then screening the mutant libraries of heavy chain CDR1, 2, 3 regions; and finally the monoclonal antibodies having high affinity are screened.

Abstract: The application belongs to the technical field of biomedicines and provides an improved anti-Vascular Endothelial Growth Factor Receptor-2 (VEGFR-2) monoclonal antibody and an application thereof. By a computer-aided simulation design, a novel phage antibody library is designed, and an improved anti-VEGFR-2 monoclonal antibody is obtained after multiple rounds of screening. Both affinity and biological activity of the antibody are higher than those of an original antibody. The antibody is capable of effectively inhibiting combination of VEGFR-2 and a ligand Vascular Endothelial Growth Factor (VEGF) thereof in vitro, and may be used in treating a tumor and a disease caused by angiogenesis such as macular degeneration.

Abstract: An art-Epidermal Growth Factor Receptor (EGFR) and anti-Cluster-of-Differentiation (CD3) bispecific antibody and uses thereof are provided. The bispecific antibody can specifically bind to antigen epidermal growth factor receptor (EGFR) on tumor cell surface and antigen CD3 molecule on immune cell surface. The single-chain antibody (ScFv) of an anti-CD3 antibody is located at the C terminal of the constant region of the anti-EGFR antibody. Further provided are a preparation method of the bispecific antibody and clinical applications of the bispecific antibody. The bispecific antibody has high affinity, and is used for treating tumor diseases caused by high expression or abnormal expression of the EGFR and other diseases caused by other overexpression of the EGFR.

Abstract: The application discloses an anti-Proprotein Convertase Subtilisin/Kexin Type 9 (PCSK9) monoclonal antibody, an amino acid sequence comprising coded antibody variable regions and Complementarity-Determining Regions (CDR), the obtaining method, and application of monoclonal antibody. The anti-PCSK9 monoclonal antibodies are screened from a phage antibody library; affinity maturation is achieved by using a method that the phage antibody library is established by means of chain displacement; After screening the light chain CDR1, 2, 3 Mutant Library of monoclonal antibodies obtained from primary screening; monoclonal antibodies having high affinity are selected; then screening the mutant libraries of heavy chain CDR1, 2, 3 regions; and finally the monoclonal antibodies having high affinity are screened.

Abstract: Provided is a method of making a fiber comprising metal nanoparticles. The method includes steps of: Step (A): providing a fiber and a metal salt aqueous solution comprising first metal ions; Step (B): making the metal salt aqueous solution contact the fiber to form a fiber containing the first metal ions; and Step (C): contacting the fiber containing the first metal ions with a second metal, and performing a reduction reaction of the first metal ions to obtain the fiber comprising metal nanoparticles, wherein the fiber comprising metal nanoparticles comprises first metal nanoparticles from a reduction of the first metal ions; wherein a standard reduction potential of the first metal ions is greater than a standard reduction potential of an ionic state of the second metal, and a difference therebetween ranges from 0.4 V to 4.0 V.