Abstract: The present disclosure provides a urea pyrolysis test device and method. The urea pyrolysis test device includes a pyrolysis reactor, a urea preparation unit and an injection unit, where the pyrolysis reactor is provided with a flue therein, the urea preparation unit is configured to prepare a urea solution, and the injection unit is connected to the urea preparation unit and includes a plurality of nozzles extending into the flue so as to spray the urea solution into the flue.

Abstract: Provided is a fly ash removal device for a flue gas, including a fly ash collection assembly adapted to be arranged in a flue. The fly ash collection assembly includes: a flow baffle, a flow guiding component, and a collecting component.

Abstract: The present invention relates to the technical field of power generation of power systems, in particular to an offshore integrated power supply system based on clean energy. The integrated power supply system comprises a power generation unit for providing energy, an energy storage unit for storing energy, a load unit for consuming energy, an energy management system, and a fuel cell, wherein the power generation unit comprises a photovoltaic power generation system, a wind power generation system, and a tidal power generation system; the energy storage unit comprises hydrogen storage and a battery pack; and the energy management system connects the power generation unit, the load unit, and the energy storage unit, and allocates the surplus energy from the power generation unit to the hydrogen storage and the battery pack after satisfying the load unit.

Abstract: Provided is a device for spraying an ammoniacal gas in an automatically-adjustable manner based on a flow rate

Abstract: Detecting fake news involves receiving a plurality of allegedly real news stories and allegedly fake news stories from one or more websites, receiving a plurality of user posts to a social media platform relating to the plurality of allegedly real news stories and allegedly fake news stories; receiving a plurality of user engagements related to the plurality of user posts, receiving user profile information, and social media network information, for users creating the plurality of user posts to the social media platform, and users participating in the engagements related to the plurality of user posts, and classifying each of the received plurality of allegedly real news stories and allegedly fake news stories as one of a real news story and a fake news story based on the analyzed content and analyzed social media context.

Abstract: This document relates to training of machine learning models. One example method involves providing a machine learning model having a first classification layer, a second classification layer, and an encoder that feeds into the first classification layer and the second classification layer. The example method also involves obtaining first training examples having explicit labels and second training examples having inferred labels. The inferred labels are based at least on actions associated with the second training examples. The example method also involves training the machine learning model using the first training examples and the second training examples using a training objective that considers first training loss of the first classification layer for the explicit labels and second training loss of the second classification layer for the inferred labels. The method also involves outputting a trained machine learning model having the encoder and the first classification layer.

Abstract: A computer-implemented framework and/or system for cyberbullying detection is disclosed. The system includes two main components: (1) A representation learning network that encodes the social media session by exploiting multi-modal features, e.g., text, network, and time; and (2) a multi-task learning network that simultaneously fits the comment inter-arrival times and estimates the bullying likelihood based on a Gaussian Mixture Model. The system jointly optimizes the parameters of both components to overcome the shortcomings of decoupled training. The system includes an unsupervised cyberbullying detection model that not only experimentally outperforms the state-of-the-art unsupervised models, but also achieves competitive performance compared to supervised models.

Abstract: This document relates to training of machine learning models. One example method involves providing a machine learning model having a first classification layer, a second classification layer, and an encoder that feeds into the first classification layer and the second classification layer. The example method also involves obtaining first training examples having explicit labels and second training examples having inferred labels. The inferred labels are based at least on actions associated with the second training examples. The example method also involves training the machine learning model using the first training examples and the second training examples using a training objective that considers first training loss of the first classification layer for the explicit labels and second training loss of the second classification layer for the inferred labels. The method also involves outputting a trained machine learning model having the encoder and the first classification layer.

Abstract: Various embodiments of systems and methods for cross media joint friend and item recommendations are disclosed herein.

Abstract: This document relates to training of machine learning models. One example method involves providing a machine learning model having a first classification layer, a second classification layer, and an encoder that feeds into the first classification layer and the second classification layer. The example method also involves obtaining first training examples having explicit labels and second training examples having inferred labels. The inferred labels are based at least on actions associated with the second training examples. The example method also involves training the machine learning model using the first training examples and the second training examples using a training objective that considers first training loss of the first classification layer for the explicit labels and second training loss of the second classification layer for the inferred labels. The method also involves outputting a trained machine learning model having the encoder and the first classification layer.

Abstract: Various embodiments of a computer-implemented system which learns textual representations while filtering out potentially personally identifying data and retaining semantic meaning within the textual representations are disclosed herein.

Abstract: Various embodiments of systems and methods for cross media joint friend and item recommendations are disclosed herein.

Abstract: Detecting fake news involves analyzing a distribution of publishers who publish many news articles, analyzing a distribution of various topics relating to the published news articles, analyzing a social media context relating to the published news articles, and detecting fake news articles among the news articles based on the analysis of the distribution of publishers, the analysis of the distribution of the various topics, and the analysis of the social media context.

Abstract: A computer-implemented framework and/or system for cyberbullying detection is disclosed. The system includes two main components: (1) A representation learning network that encodes the social media session by exploiting multi-modal features, e.g., text, network, and time; and (2) a multi-task learning network that simultaneously fits the comment inter-arrival times and estimates the bullying likelihood based on a Gaussian Mixture Model. The system jointly optimizes the parameters of both components to overcome the shortcomings of decoupled training. The system includes an unsupervised cyberbullying detection model that not only experimentally outperforms the state-of-the-art unsupervised models, but also achieves competitive performance compared to supervised models.

Abstract: A news article may include sentences and have associated comments. A embodiment determines semantic correlation between each sentence and each comment to generate correlation degrees between the sentences and the comments, determines sentence attention weights of the sentences and comment attention weights of the comments based on the correlation degrees, and detect whether the news article is fake based on latent representations of the sentences and the comments, the sentence attention weights and the comment attention weights. A list of sentences and a list of comments may be selected based on the sentence attention weights and the comment attention weights, respectively, to provide explanation for a detection result.

Abstract: This document relates to training of machine learning models. One example method involves providing a machine learning model having a first classification layer, a second classification layer, and an encoder that feeds into the first classification layer and the second classification layer. The example method also involves obtaining first training examples having explicit labels and second training examples having inferred labels. The inferred labels are based at least on actions associated with the second training examples. The example method also involves training the machine learning model using the first training examples and the second training examples using a training objective that considers first training loss of the first classification layer for the explicit labels and second training loss of the second classification layer for the inferred labels. The method also involves outputting a trained machine learning model having the encoder and the first classification layer.

Abstract: Various embodiments of a computer-implemented system which learns textual representations while filtering out potentially personally identifying data and retaining semantic meaning within the textual representations are disclosed herein.

Abstract: Detecting fake news involves receiving a plurality of allegedly real news stories and allegedly fake news stories from one or more websites, receiving a plurality of user posts to a social media platform relating to the plurality of allegedly real news stories and allegedly fake news stories; receiving a plurality of user engagements related to the plurality of user posts, receiving user profile information, and social media network information, for users creating the plurality of user posts to the social media platform, and users participating in the engagements related to the plurality of user posts, and classifying each of the received plurality of allegedly real news stories and allegedly fake news stories as one of a real news story and a fake news story based on the analyzed content and analyzed social media context.

Abstract: Various embodiments of systems and methods for cross media joint friend and item recommendations are disclosed herein.

Abstract: Detecting fake news involves analyzing a distribution of publishers who publish many news articles, analyzing a distribution of various topics relating to the published news articles, analyzing a social media context relating to the published news articles, and detecting fake news articles among the news articles based on the analysis of the distribution of publishers, the analysis of the distribution of the various topics, and the analysis of the social media context.