Ricoh Company, Ltd.

Coupang, Corporation

Honeywell International Inc.

International Business Machines Corporation

Huawei Technologies Co., Limited

Beijing Kingsoft Cloud Network Technology Co., Ltd.

Beijing Kingsoft Cloud Technology Co., Ltd.

Bin Dong

Shanshan Jiang

Yixuan Tong

Lei Ding

Junzhao Zhang

Chuanchuan Han

Yongwei Zhang

Zhan Chen

Zhiyong Huang

Jiashi Zhang

Kyusung Kim

Qingqiu Ginger Shao

Jichuan Zheng

Rui Cheng

Gary Kumagai

Liang Liang

Roy Araki

Eric Shi

Wenyan Guo

Yihan Li

Yang Liu

Yue Ma

Tie Li

Zhong Hai Chen

Guanglei Zhang

Xu Guang Gu

Wenqiang Wang

Mohit Singhai

Andrew Earl

Tianxiong Xiao

Lu Luo

Jinkun Liu

Boyan Liu

Qi Li

A method and an apparatus for named entity recognition, and a non-transitory computer-readable recording medium are provided. In the method, text elements are traversed according to a text span to obtain candidate entity words. Then, a class to which the candidate entity word belongs is recognized. The recognizing of the class includes generating a prompt template corresponding to the candidate entity word, and concatenating the text to be recognized and the prompt template to obtain a concatenated text; generating vector representations of the text elements in the concatenated text; generating the vector representation of the candidate entity word according to the vector representations of the text elements of each candidate entity word in the concatenated text, and the vector representation of the text element of the mask word; and classifying the vector representation of the candidate entity word to obtain the class of the candidate entity word.

Method and apparatus for named entity recognition, and non-transitory computer-readable recording medium

A method for deploying a service includes a controller that creates a service instance corresponding to a service, and creates, in the service instance, a first network instance and a second network instance that correspond to the service. The first network instance is a network instance corresponding to a first device, and the first network instance includes a device identifier of the first device. The second network instance is a network instance corresponding to a first cloud provider edge PE, and the second network instance includes a device identifier of the first cloud PE. The first cloud PE is for connecting to the first cloud.

Method, apparatus, and system for deploying service

Disclosed are a table recognition method and apparatus. The table recognition method includes steps of obtaining an image vision feature and a character content feature of a table image; fusing the image vision feature and the character content feature of the table image to acquire a first fusion feature, and carrying out recognition based on the first fusion feature to acquire a table structure; and performing, based on the table structure, character recognition on the table image to acquire table character contents.

Table recognition method and apparatus and non-transitory computer-readable medium

A system for database backup including a controller, a metadata database coupled to the controller, a backup tool coupled to the controller, and a plurality of nodes coupled to the backup tool and accessed through a file system. The controller is configured to determine a first data change in node; in response to determining the first data change, freeze the file system; and generate an incremental snapshot of the at least one of the plurality of nodes (where the incremental snapshot including a portion of the at least one of the plurality of nodes). The controller is also configured to upload an encrypted copy of the incremental snapshot to a cloud storage; update the metadata database by storing an entry associating the incremental snapshot with the at least one of the plurality of nodes and unfreezing the file system.

Cloud-based database backup and recovery

Disclosed are a semantic matching and retrieval method and apparatus. The semantic matching and retrieval method includes steps of obtaining both the vector representation of a query text and the vector representation of a document text; obtaining the final vector representation of the query text; obtaining the final vector representation of the document text; calculating, based on the final vector representation of the query text and the final vector representation of the document text, the similarity score between the query text and the document text; and selecting, based on the similarity scores between the query text and a plurality of document texts, a document text matching the query text from the plurality of document texts.

Semantic matching and retrieval method and apparatus and non-transitory computer-readable medium

A method and an apparatus for machine reading comprehension, and a non-transitory computer-readable recording medium are provided. In the method, a paragraph-question pair is obtained, and subword vectors corresponding to subwords in the paragraph-question pair are generated. Then, for each subword, relative positions of the subword with respect to the other subwords are determined based on distances, and self-attention information of the subword in a first part and mutual attention information of the subword in a second part are calculated by using the relative positions and the subword vector. Then, a fusion vector of the subword is generated based on the self-attention information and the mutual attention information. Then, the fusion vectors of the subwords are input to a decoder of a machine reading comprehension model so as to obtain an answer predicted by the decoder.

Method and apparatus for machine reading comprehension, and non-transitory computer-readable recording medium

A computer-implemented system and method for generating a data transaction log is disclosed. The method has the steps of: generating a registry; maintaining at least one data source agent, each data source agent being configured to: query the registry to receive an address of the one or more first data table; receiving, in response to the query, the at least one transaction logs from at least one first storage, wherein each transaction log is associated with a transaction time; compiling one or more modified transaction logs; and maintaining at least one streaming process, each streaming process being configured to: generate the at least one second data table based on the modified transaction logs for storage in the at least one second storage; and generate display based on the at least one second data table.

Systems and methods for multi-nodal stream processing framework for partitioned database

Methods and systems for database query efficiency improvement are disclosed. In one embodiment, a method includes mirroring a primary database to a secondary database; creating a testing database comprising the schema; receiving a query; running the query on the testing database; and evaluating the query by: identifying predicates in the query; determining most common values for each column name by querying the secondary database; creating, for each column name, a list comprising at least one of the most common values; creating a test predicate comprising one of the column names and an entry for the list corresponding to the column name; creating a test query comprising one or more test predicates; determining a resource utilization of the query by running each of the test queries on the secondary database; and providing, to a user interface for display, an efficiency improvement recommendation when the resource utilization exceeds a threshold.

Systems and methods for database query efficiency improvement

A method and an apparatus for sequence labeling on an entity text, and a non-transitory computer-readable recording medium are provided. In the method, a start position of an entity text within a target text is determined. Then, a first matrix is generated based on the start position of the entity text. Elements in the first matrix indicates focusable weights of each word with respect to other words in the target text. Then, a named entity recognition model is generated using the first matrix. The named entity recognition model is obtained by training using first training data, the first training data includes word embeddings corresponding to respective texts in a training text set, and the texts are texts whose entity label has been labeled. Then, the target text is input to the named entity recognition model, and probability distribution of the entity label is output.

Method and apparatus for sequence labeling on entity text, and non-transitory computer-readable recording medium

An example system for determining a performance status of an environmental control system (ECS) in a vehicle includes memory and processing circuitry. The processing circuitry is configured to determine, based on one or more of the aircraft data, the weather data, or the trending data, an estimated compressor exit temperature. The processing circuitry is configured to determine a current residual compressor exit temperature based on the estimated compressor exit temperature and a current compressor exit temperature. The processing circuitry is configured to determine whether a residual condition is met based at least in part on the residual compressor exit temperature and whether a pack condition is met based at least in part on a pack temperature. The processing circuitry is configured to provide an indication of the performance status of the ECS based on at least one of the residual condition or the pack condition being met.

Environmental control system performance monitoring

Systems and methods for managing a highly available distributed hybrid database comprising: a memory storing instructions; and one or more processors configured to execute the instructions to: receive a query from a user device to retrieve data from a distributed database comprising a source node, a first plurality of replica nodes, and a second plurality of replica nodes, wherein the source node and the first plurality of replica nodes form a transactional cluster, and wherein the second plurality of replica nodes forms an analytical cluster; determine whether to process the query using the transactional cluster or the analytical cluster based on one or more rules; translate the query into a first protocol that the determined cluster comprehends; select a replica node corresponding to the determined cluster; process the query using the selected replica node; and send data associated with results from processing the query to the user device.

Systems and methods for managing a highly available distributed hybrid transactional and analytical database

According to certain aspects of the disclosure, a computer-implemented method may be used for detecting health status of an environmental control system. The method may include receiving aircraft data of an aircraft and receiving flight data of an aircraft. Calculating a predicted performance of the aircraft based on the received aircraft data and the received flight data and generating at least one model scalar or residual, wherein the at least one model scalar or residual is generated based on the aircraft data of the aircraft. Identifying at least one pattern from the at least one model scalar or residual and classifying the at least one pattern into at least one of a plurality of classifications. Identifying a failure of modes or components from the classifications and transmitting a maintenance report once the failure of modes or components is identified.

Pattern classification system with smart data collection for environmental control system fault isolation

Systems and methods for managing a highly available distributed database comprising: a memory storing instructions; and one or more processors configured to execute the instructions to: determine that a source node, in a distributed database comprising the source node and one or more replica nodes, is not available; select a most-updated replica node from the one or more replica nodes; switch a role of the most-updated replica node to source; update a data store to label the source node as unavailable and the selected replica node as being a promoted source node; send a notification to a user device to update a database topology based on the updated data store; determine whether the user device has updated the database topology; and upon determining the user device has not updated the database topology, continue to send the notification to the user device until the user device has updated the database topology.

Systems and methods for managing a highly available and scalable distributed database in a cloud computing environment

A method and an apparatus for fusing position information, and a non-transitory computer-readable recording medium are provided. In the method, words of an input sentence are segmented to obtain a first sequence of words in the input sentence, and absolute position information of the words in the first sequence is generated. Then, subwords of the words in the first sequence are segmented to obtain a second sequence including subwords, and position information of the subwords in the second sequence are generated, based on the absolute position information of the words in the first sequence, to which the respective subwords belong. Then, the position information of the subwords in the second sequence are fused into a self-attention model to perform model training or model prediction.

Method and apparatus for fusing position information, and non-transitory computer-readable recording medium

A method and an apparatus for recognizing an intention, and a non-transitory computer-readable recording medium are provided. The method includes learning vectors of knowledge base elements in corpus samples, and converting the corpus samples into row vectors composed of the vectors of the knowledge base elements in a knowledge base; extracting feature vectors from respective pooling windows in the corpus samples by hierarchical pooling, determining weights positively correlated with similarities between texts within the respective pooling windows and the respective corpus samples, performing weighting on the extracted feature vectors to obtain feature vectors of the respective pooling windows, and obtaining feature vectors of the respective corpus samples composed of the feature vectors of the pooling windows; training a vector-based intention recognition classifier, based on the feature vectors of the corpus samples; and recognizing an intention in querying a corpus, using the trained intention recognition classifier.

Growing community of inventors

Beijing, China

Bin Dong