Sang N Song

US Patent:

20130151443, Jun 13, 2013

Filed:

Oct 3, 2012

Appl. No.:

13/573727

Inventors:

AOL Inc. - Dulles VA, US
Sang Chul Song - Aldie VA, US
Vineet Mahajan - Ashburn VA, US
Elena Haliczer - New York NY, US

Assignee:

AOL Inc. - Dulles VA

International Classification:

G06N 99/00

US Classification:

706 12

Abstract:

Computerized systems and methods are disclosed for performing contextual classification of objects using supervised and unsupervised training. In accordance with one implementation, content reviewers may review training objects and submit supervised training data for preprocessing and analysis. The supervised training data may be preprocessed to identify key terms and phrases, such as by stemming, tokenization, or n-gram analysis, and form vectorized objects. The vectorized objects may be used to train one or more models for subsequent classification of objects. In certain implementations, preprocessing or training, among other steps, may be performed in parallel over multiple machines to improve efficiency. The disclosed systems and methods may be used in a wide variety of applications, such as article classification and content moderation.

US Patent:

20230100813, Mar 30, 2023

Filed:

Dec 7, 2022

Appl. No.:

18/062797

Inventors:

- Dulles VA, US
Sang Chul SONG - Aldie VA, US
Sean Christopher TIMM - Herndon VA, US
Thu Rein KYAW - Reston VA, US
Saad MUFTI - Fairfax VA, US

International Classification:

G06Q 30/02

Abstract:

Systems and methods are disclosed for online distribution of content based on a user sentiment index. The method may include receiving, over a network and from a user device, one or more user generated inputs and calculating the user sentiment index based on the one or more user generated inputs. The method may also include receiving, over the network, from a content or advertising provider, instructions on publishing content or advertising to a webpage based on the calculated user sentiment index, and publishing content for display on user devices over the network based on a comparison of the calculated user sentiment index and the received instructions.

US Patent:

20210365300, Nov 25, 2021

Filed:

Nov 21, 2018

Appl. No.:

16/198133

Inventors:

Thu R. KYAW - Reston VA, US
Jonathan JI - Aldie VA, US
Saad MUFTI - Fairfax VA, US
Sudhir ACHUTHAN - Vienna VA, US
Sang Chul SONG - Aldie VA, US

International Classification:

G06F 9/50
G06F 17/30

Abstract:

Methods, systems, and computer-readable media are disclosed for dynamic partitioning in distributed computing environments. One method includes: receiving a first data set and a second data set; mapping the first data set into a first set of key-value pairs; mapping the second data set into a second set of key-value pairs; estimating, using a sketch, a frequency count for each key based on the first set of key-value pairs and the second set of key-value pairs; determining whether the estimated frequency count for each key is greater than or equal to a predetermined threshold; and partitioning the key when the estimated frequency count for the key is greater than or equal to the predetermined threshold.

US Patent:

20210103478, Apr 8, 2021

Filed:

Dec 15, 2020

Appl. No.:

17/122849

Inventors:

- New York NY, US
Jonathan JI - Aldie VA, US
Saad MUFTI - Fairfax VA, US
Sudhir ACHUTHAN - Vienna VA, US
Sang Chul SONG - Aldie VA, US

International Classification:

G06F 9/50
G06F 16/28

Abstract:

Methods, systems, and computer-readable media are disclosed for dynamic partitioning in distributed computing environments. One method includes: receiving a first data set and a second data set; mapping the first data set into a first set of key-value pairs; mapping the second data set into a second set of key-value pairs; estimating, using a sketch, a frequency count for each key based on the first set of key-value pairs and the second set of key-value pairs; determining whether the estimated frequency count for each key is greater than or equal to a predetermined threshold; and partitioning the key when the estimated frequency count for the key is greater than or equal to the predetermined threshold.

US Patent:

20210012385, Jan 14, 2021

Filed:

Sep 30, 2020

Appl. No.:

17/038378

Inventors:

- Dulles VA, US
Sang Chul SONG - Aldie VA, US
Sean Christopher TIMM - Herndon VA, US
Thu Rein KYAW - Reston VA, US
Saad MUFTI - Fairfax VA, US

International Classification:

G06Q 30/02

Abstract:

Systems and methods are disclosed for online distribution of content based on a user sentiment index. The method may include receiving, over a network and from a user device, one or more user generated inputs and calculating the user sentiment index based on the one or more user generated inputs. The method may also include receiving, over the network, from a content or advertising provider, instructions on publishing content or advertising to a webpage based on the calculated user sentiment index, and publishing content for display on user devices over the network based on a comparison of the calculated user sentiment index and the received instructions.

US Patent:

20200143289, May 7, 2020

Filed:

Jan 8, 2020

Appl. No.:

16/737581

Inventors:

- Dulles VA, US
Sang Chul SONG - Aldie VA, US
Vineet MAHAJAN - Ashburn VA, US
Elena HALICZER - New York NY, US

International Classification:

G06N 20/00
G06N 5/02
G06F 40/30
G06F 40/20

Abstract:

Computerized systems and methods are disclosed for performing contextual classification of objects using supervised and unsupervised training. In accordance with one implementation, content reviewers may review training objects and submit supervised training data for preprocessing and analysis. The supervised training data may be preprocessed to identify key terms and phrases, such as by stemming, tokenization, or n-gram analysis, and form vectorized objects. The vectorized objects may be used to train one or more models for subsequent classification of objects. In certain implementations, preprocessing or training, among other steps, may be performed in parallel over multiple machines to improve efficiency. The disclosed systems and methods may be used in a wide variety of applications, such as article classification and content moderation.

US Patent:

20170371892, Dec 28, 2017

Filed:

Jun 22, 2016

Appl. No.:

15/189158

Inventors:

- New York NY, US
Jonathan JI - Aldie VA, US
Saad MUFTI - Fairfax VA, US
Sudhir ACHUTHAN - Vienna VA, US
Sang Chul SONG - Aldie VA, US

International Classification:

G06F 17/30

Abstract:

Methods, systems, and computer-readable media are disclosed for dynamic partitioning in distributed computing environments. One method includes: receiving a first data set and a second data set; mapping the first data set into a first set of key-value pairs; mapping the second data set into a second set of key-value pairs; estimating, using a sketch, a frequency count for each key based on the first set of key-value pairs and the second set of key-value pairs; determining whether the estimated frequency count for each key is greater than or equal to a predetermined threshold; and partitioning the key when the estimated frequency count for the key is greater than or equal to the predetermined threshold.

US Patent:

20150310352, Oct 29, 2015

Filed:

Jul 6, 2015

Appl. No.:

14/792585

Inventors:

- Dulles VA, US
Sang Chul SONG - Aldie VA, US
Vineet MAHAJAN - Ashburn VA, US
Elena HALICZER - New York NY, US

International Classification:

G06N 99/00
G06N 5/02

Abstract:

Computerized systems and methods are disclosed for performing contextual classification of objects using supervised and unsupervised training. In accordance with one implementation, content reviewers may review training objects and submit supervised training data for preprocessing and analysis. The supervised training data may be preprocessed to identify key terms and phrases, such as by stemming, tokenization, or n-gram analysis, and form vectorized objects. The vectorized objects may be used to train one or more models for subsequent classification of objects. In certain implementations, preprocessing or training, among other steps, may be performed in parallel over multiple machines to improve efficiency. The disclosed systems and methods may be used in a wide variety of applications, such as article classification and content moderation.