Patents by Inventor Chang Hua Sun
Chang Hua Sun has filed for patents to protect the following inventions. This listing includes patent applications that are pending as well as patents that have already been granted by the United States Patent and Trademark Office (USPTO).
-
Patent number: 11501083Abstract: Techniques are provided for training, by a system operatively coupled to a processor, an attention weighted recurrent neural network encoder-decoder (AWRNNED) using an iterative process based on one or more paragraphs of agent sentences from respective transcripts of one or more conversations between one or more agents and one or more customers, and based on one or more customer response sentences from the respective transcripts, and generating, by the system, one or more groups respectively comprising one or more agent sentences and one or more customer response sentences selected based on attention weights of the AWRNNED.Type: GrantFiled: December 24, 2020Date of Patent: November 15, 2022Assignee: INTERNATIONAL BUSINESS MACHINES CORPORATIONInventors: Ke Ke Cai, Jing Ding, Zhong Su, Chang Hua Sun, Li Zhang, Shi Wan Zhao
-
Publication number: 20210117626Abstract: Techniques are provided for training, by a system operatively coupled to a processor, an attention weighted recurrent neural network encoder-decoder (AWRNNED) using an iterative process based on one or more paragraphs of agent sentences from respective transcripts of one or more conversations between one or more agents and one or more customers, and based on one or more customer response sentences from the respective transcripts, and generating, by the system, one or more groups respectively comprising one or more agent sentences and one or more customer response sentences selected based on attention weights of the AWRNNED.Type: ApplicationFiled: December 24, 2020Publication date: April 22, 2021Inventors: Ke Ke Cai, Jing Ding, Zhong Su, Chang Hua Sun, Li Zhang, Shi Wan Zhao
-
Patent number: 10902205Abstract: Techniques are provided for training, by a system operatively coupled to a processor, an attention weighted recurrent neural network encoder-decoder (AWRNNED) using an iterative process based on one or more paragraphs of agent sentences from respective transcripts of one or more conversations between one or more agents and one or more customers, and based on one or more customer response sentences from the respective transcripts, and generating, by the system, one or more groups respectively comprising one or more agent sentences and one or more customer response sentences selected based on attention weights of the AWRNNED.Type: GrantFiled: October 25, 2017Date of Patent: January 26, 2021Assignee: INTERNATIONAL BUSINESS MACHINES CORPORATIONInventors: Ke Ke Cai, Jing Ding, Zhong Su, Chang Hua Sun, Li Zhang, Shi Wan Zhao
-
Patent number: 10671686Abstract: A method, apparatus and system for processing webpage data. The method includes: in response to a webpage being opened, sending a link contained in the webpage to a network side device; receiving a group identification from the network side device, the group identification being determined by the network side device according to the link and used to specify a group the link belongs to; determining whether there is a browsed link belonging to the group specified by the group identification; and in response to determining there is a browsed link belonging to the group specified by the group identification, prompting that webpage content pointed by the link contained in the webpage has been browsed.Type: GrantFiled: February 24, 2014Date of Patent: June 2, 2020Assignee: INTERNATIONAL BUSINESS MACHINES CORPORATIONInventors: Sheng Hua Bao, Dong Xu Duan, Wei Hong Qian, Chang Hua Sun
-
Patent number: 10572585Abstract: This disclosure provides a computer-implemented method. The method may include extracting one or more features based on a first utterance from a first interlocutor in a dialog and a second utterance from a second interlocutor in the dialog. The method may further include inferring one or more personality traits of the first interlocutor based on the one or more extracted features from the dialog.Type: GrantFiled: November 30, 2017Date of Patent: February 25, 2020Assignee: International Business Machines CoporationInventors: En Liang Xu, Chang Hua Sun, Shi Wan Zhao, Ke Ke Cai, Yue Chen, Li Zhang, Zhong Su
-
Patent number: 10389677Abstract: Embodiments of the invention provide a computer-implemented method, computing system and computer program product for analyzing a message in a social network. The method comprises identifying an entity from the message; detecting historical popularity of the entity in a social network; identifying a topic from the message; detecting historical popularity of the topic in the social network; and generating an entity-topic correlation factor for the entity and the topic based on the historical popularity of the entity and the historical popularity of the topic. Results obtained with embodiments of the invention may be provided to popularity prediction tools for improving popularity prediction of messages in social networks.Type: GrantFiled: December 23, 2016Date of Patent: August 20, 2019Assignee: International Business Machines CorporationInventors: Ke Ke Cai, Hong Lei Guo, Jian Min Jiang, Zhong Su, Chang Hua Sun, Guo Yu Tang
-
Publication number: 20190163735Abstract: This disclosure provides a computer-implemented method. The method may include generating one or more features based on a first utterance from a first interlocutor in a dialog and a second utterance from a second interlocutor in the dialog.Type: ApplicationFiled: November 30, 2017Publication date: May 30, 2019Inventors: En Liang XU, Chang Hua SUN, Shi Wan ZHAO, Ke Ke CAI, Yue CHEN, Li ZHANG, Zhong SU
-
Publication number: 20190121853Abstract: Techniques are provided for training, by a system operatively coupled to a processor, an attention weighted recurrent neural network encoder-decoder (AWRNNED) using an iterative process based on one or more paragraphs of agent sentences from respective transcripts of one or more conversations between one or more agents and one or more customers, and based on one or more customer response sentences from the respective transcripts, and generating, by the system, one or more groups respectively comprising one or more agent sentences and one or more customer response sentences selected based on attention weights of the AWRNNED.Type: ApplicationFiled: October 25, 2017Publication date: April 25, 2019Inventors: Ke Ke Cai, Jing Ding, Zhong Su, Chang Hua Sun, Li Zhang, Shi Wan Zhao
-
Publication number: 20180183747Abstract: Embodiments of the invention provide a computer-implemented method, computing system and computer program product for analyzing a message in a social network. The method comprises identifying an entity from the message; detecting historical popularity of the entity in a social network; identifying a topic from the message; detecting historical popularity of the topic in the social network; and generating an entity-topic correlation factor for the entity and the topic based on the historical popularity of the entity and the historical popularity of the topic. Results obtained with embodiments of the invention may be provided to popularity prediction tools for improving popularity prediction of messages in social networks.Type: ApplicationFiled: December 23, 2016Publication date: June 28, 2018Inventors: Ke Ke CAI, Hong Lei GUO, Jian Min JIANG, Zhong SU, Chang Hua SUN, Guo Yu TANG
-
Publication number: 20170011419Abstract: A mechanism is provided for personalizing a user's E-commerce environment. Identified lifecycle state transactions associated with the user are modeled by performing a lifecycle state transition probability calculation utilizing collected social media data and transaction data. Utilizing the model of the identified lifecycle state transactions, a two-level Hidden Markov Model (HMM) lifecycle model is generated for current lifecycle states being experienced by the user. Utilizing the two-level HMM lifecycle model for current lifecycle states being experienced by the user, one or more future behavioral predictions are generated with regard to the user's lifecycle. One or more E-commerce recommendations are then issued to the user based on the one or more future behavioral predictions.Type: ApplicationFiled: July 9, 2015Publication date: January 12, 2017Inventors: Ke Ke Cai, Dong Xu Duan, Zhong Su, Chang Hua Sun, Shi Lei Zhang, Shi Wan Zhao
-
Patent number: 9003433Abstract: Business application logic corresponding to a business is generated according to requirements of the business. The business application logic includes Web services and Widgets. The Widgets in the business application logic are mashed up to create a Widget mashup application. An adaptation relationship is established between the Web services and the Widgets in the Widget mashup application. An executable business process is generated for the business based on the business application logic and the adaptation relationship. An executable mashup Widget application is generated based on the Widget mashup application and the adaptation relationship.Type: GrantFiled: January 23, 2012Date of Patent: April 7, 2015Assignee: International Business Machines CorporationInventors: Shuang Liang, Xiao Xing Liang, Xin Peng Liu, Chang Hua Sun, Xi Ning Wang, Liang Xue, Yu Chen Zhou
-
Publication number: 20140244646Abstract: A method, apparatus and system for processing webpage data. The method includes: in response to a webpage being opened, sending a link contained in the webpage to a network side device; receiving a group identification from the network side device, the group identification being determined by the network side device according to the link and used to specify a group the link belongs to; determining whether there is a browsed link belonging to the group specified by the group identification; and in response to determining there is a browsed link belonging to the group specified by the group identification, prompting that webpage content pointed by the link contained in the webpage has been browsed.Type: ApplicationFiled: February 24, 2014Publication date: August 28, 2014Applicant: International Business Machines CorporationInventors: Sheng Hua Bao, Dong Xu Duan, Wei Hong Qian, Chang Hua Sun
-
Publication number: 20120198481Abstract: Business application logic corresponding to a business is generated according to requirements of the business. The business application logic includes Web services and Widgets. The Widgets in the business application logic are mashed up to create a Widget mashup application. An adaptation relationship is established between the Web services and the Widgets in the Widget mashup application. An executable business process is generated for the business based on the business application logic and the adaptation relationship. An executable mashup Widget application is generated based on the Widget mashup application and the adaptation relationship.Type: ApplicationFiled: January 23, 2012Publication date: August 2, 2012Applicant: INTERNATIONAL BUSINESS MACHINES CORPORATIONInventors: Shuang Liang, Xiao Xing Liang, Xin Peng Liu, Chang Hua Sun, Xi Ning Wang, Liang Xue, Yu Chen Zhou