International Business Machines Corporation

Cccc Highway Consultants Co. Ltd.

China Communications Construction Company Limited

Hai Ji

Yuan Yuan Li

Xiaoyang Yang

Chun Guang Zeng

Zi Jian Ji

Xiao Yang

Kun Yan Yin

William J Carpenter

Wen Bo Ma

Chun Guang Zheng

Jia Mi

Qiang Wang

Xin Peng Liu

Xiaodong Wang

Wen Jie Wang

Wei Lin

Zhong Fang Yuan

Yao Liang Chen

Tong Liu

Xiaodong Liu

Zhigang Zhang

He Li

Tao Duan

De Shuo Kong

Yuan Lin Yang

Hai B Zou

Ming Lin

Ke Deng

Qian Cheng

Zheng Jie

Rui W W Wang

Haiqing Yin

Jibing Gao

Lingfeng Liu

Qi Y Wang

Huaiping Su

Zhu Lin Zhang

Yonggang Lv

A method includes: receiving, by a computer device, a user question; extracting, by the computer device and from a knowledge base, multiple answers to the user question; detecting, by the computer device, a first relationship entity that exists in a subset of the multiple answers to the user question; determining, by the computer device, that the first relationship entity is a condition, the condition being a relationship entity that has different values in the subset of the multiple answers; generating, by the computer device, a first follow-up question that solicits a value of the condition; receiving, by the computer device, an answer to the first follow-up question; and determining, by the computer device, a number of the subset of the multiple answers that have a value of the condition that matches the answer to the first follow-up question.

Intelligent leading multi-round interactive automated information system

A method, system and computer program product for speech recognition using multiple languages includes receiving, by one or more processors, an input from a user, the input includes a sentence in a first language. The one or more processors translate the sentence to a plurality of languages different than the first language, and create vectors associated with the plurality of languages, each vector includes a representation of the sentence in each of the plurality of languages. The one or more processors calculate eigenvectors for each vector associated with a language in the plurality of languages, and based on the calculated eigenvectors, a score is assigned to each of the plurality of languages according to a relevance for determining a meaning of the sentence.

Cognitive analysis for speech recognition using multi-language vector representations

A plurality of documents is received. From one or more documents within the plurality, a set of image features is extracted. A document feature repository is generated by storing one or more sets of image features, by document. A document search query is received. The document search query is pre-processed to generate a set of variable document search queries. The document feature repository is searched using the set of variable document search queries. The search results are presented to a user.

Document feature repository management

A latest status record of an object in a local storage node is obtained as a local status record of the object. The status record includes a name of an operation changing the object, a timestamp of the operation changing the object, a first status flag, and a second status flag. The first status flag indicates whether a synchronization of the object from current object storage node to another object storage node is executed. The second status flag indicates whether a synchronization of the object from another object storage node to the current object storage node is executed. The latest status record of the object is obtained of the object in from a remote object storage node as a remote status record of the object. A synchronization of the object is executed based on the local status record and the remote status record.

Synchronizing object in local object storage node

Automated multi-round task-based conversation with a user is disclosed. In a first round conversation between a user and a data processing system involving first task(s), the data processing system cognitively identifies first entity(ies) and one or more first intent(s), based on a first input statement received by the data processing system from the user. In a subsequent second round conversation between the user and the data processing system involving second task(s) different from the first task(s), the data processing system cognitively identifies second entity(ies) and second intent(s), based on a second input statement received by the data processing system from the user. The data processing system reuses at least one of the first entity(ies) in the second round conversation.

Reusing entities in automated task-based multi-round conversation

A method, computer system, and a computer program product for detecting a session status based on a cookie associated with the session is provided. The present invention may include receiving an access request to a specified location associated with a server computer. The present invention may also include determining that the received access request has the cookie corresponding with the specified location. The present invention may also include receiving a last refresh time from the cookie. The present invention may then include determining the session status based on the retrieved last refresh time, a current request time, a refresh interval, and an overdue value.

Cookie based session timeout detection and management

An example operation may include one or more of identifying partitions associated with a blockchain, identifying a new transaction to be committed to the blockchain, determining a partition code assigned to the new transaction, and committing the new transaction to one of the partitions associated with the partition code.

Partitioning of a blockchain ledger

The present application discloses a final joint of an immersed tunnel, a prefabrication method and an installation method, wherein the final joint includes two end surfaces connected with installed adjacent tube sections; the two end surfaces are both tilted surfaces, so that the longitudinal profile of the final joint along an installation direction is of an inverted trapezoid structure; and the final joint further may be of a structure with a tube section I and a tube section II which are connected with each other. The final joint of the immersed tunnel is simple in structure, convenient to control and relatively high in precision, thereby reducing lots of open sea diving work and lowering a risk of installation quality defects; as prefabrication procedures are simple, the final joint may be prefabricated in a land factory and then transported to the site, thereby reducing influence of weather conditions on construction; a body structure of the final joint is prefabricated in the factory, and then the overall final joint is transported to the site for installation; water stop systems realize quick water stop, thus forming a dry construction environment; and therefore, the influence of weather and tidal current conditions on a project may be reduced, and a quality risk may be lowered.

Final joint of immersed tunnel as well as prefabrication method and installation method

A method and system for storage of an aggregation file and method and system for compression of the same. The method for compressing an aggregation file includes: acquiring the aggregation file to be compressed; copying remaining files in the acquired aggregation file into a new aggregation file based on metadata of a deleted object stored in a deletion file corresponding to the acquired aggregation file; and removing the acquired aggregation file. The present invention also provides a system for compressing an aggregation file and a method and system for storing an aggregation file.

Storage and compression of an aggregation file

A method and an apparatus for object storage are disclosed. The method includes, in response to an update to metadata of an object in an object storage region to produce updated metadata of the object, recording a metadata object associated with the object in a metadata change storage region separated from the object storage region. The metadata object includes the updated metadata of the object and pointer information. The pointer information points to a first storage location representing a storage location of the object in the object storage region. The method also includes modifying index information associated with the object to produce modified index information. The modified index information indicates the update to the metadata and a second storage location representing a storage location of the metadata object in the metadata change storage region.

Method and apparatus for object storage

A method, system, and computer program product for detecting consistency between first and second data replicas is provided. Detecting consistency between data replicas includes acquiring first characteristic information of a first data replica and second characteristic information of a second data replica, the first and second characteristic information being associated with keys of the respective first and second data replicas. Detecting consistency between data replicas further include determining, based on a predetermined threshold value of a number of determination times, whether the first characteristic information is identical to the second characteristic information of the second data replica. In response to the determining that the first characteristic information and the second characteristic information are identical, whether the first data replica is consistent with the second data replica may then be confirmed.

Detection of data replication consistency

Growing community of inventors

Beijing, China

Hai Ji