Sumit R Taank

7085771, Aug 1, 2006

May 17, 2002

10/150795

Christina Yip Chung - Sunnyvale CA, US
Jinhui Liu - Sunnyvale CA, US
Alpha Luk - San Jose CA, US
Jianchang Mao - San Jose CA, US
Sumit Taank - Austin TX, US
Vamsi Vutukuru - Austin TX, US

Verity, Inc - Sunnyvale CA

L06F 17/30

707102, 707 7

The invention is a method, system and computer program for automatically discovering concepts from a corpus of documents and automatically generating a labeled concept hierarchy. The method involves extraction of signatures from the corpus of documents. The similarity between signatures is computed using a statistical measure. The frequency distribution of signatures is refined to alleviate any inaccuracy in the similarity measure. The signatures are also disambiguated to address the polysemy problem. The similarity measure is recomputed based on the refined frequency distribution and disambiguated signatures. The recomputed similarity measure reflects actual similarity between signatures. The recomputed similarity measure is then used for clustering related signatures. The signatures are clustered to generate concepts and concepts are arranged in a concept hierarchy. The concept hierarchy automatically generates query for a particular concept and retrieves relevant documents associated with the concept.

20030177000, Sep 18, 2003

Mar 12, 2002

10/096048

Jianchang Mao - San Jose CA, US
Sumit Taank - Austin TX, US
Christina Chung - Sunnyvale CA, US
Alpha Luk - San Jose CA, US

Verity, Inc.

G06F017/27

704/009000

The present invention provides a method, system and computer program for naming a cluster, or a hierarchy of clusters, of words and phrases that have been extracted from a set of documents. The invention takes these clusters as the input and generates appropriate labels for the clusters using a lexical database. Naming involves first finding out all possible word senses for all the words in the cluster, using the lexical database; and then augmenting each word sense with words that are semantically similar to that word sense to form respective definition vectors. Thereafter, word sense disambiguation is done to find out the most relevant sense for each word. Definition vectors are clustered into groups. Each group represents a concept. These concepts are thereafter ranked based on their support. Finally, a pre-specified number of words and phrases from the definition vectors of the dominant concepts are selected as labels, based on their generality in the lexical database. Therefore, the labels may not necessarily consist of the original words in the cluster. A hierarchy of clusters is named in a recursive fashion starting from leaf clusters. Dominant concepts in child clusters are propagated into their parent to reduce the labeling complexity of parent clusters.

20170220577, Aug 3, 2017

Jan 29, 2016

15/011221

- Menlo Park CA, US
Sourav Chatterji - Fremont CA, US
Sumit Taank - San Jose CA, US
Deepayan Chakrabarti - Austin TX, US

G06F 17/30

Systems, methods, and non-transitory computer-readable media can determine one or more respective topics of interest for at least some users of a social networking system. At least some of the topics can be propagated to at least a first user, wherein the propagated topics were determined to be of interest to users that follow the first user in the social networking system. At least one topic from the propagated topics for which the first user is a topical authority is determined.

20140310270, Oct 16, 2014

Apr 16, 2013

13/864046

- Bentonville AR, US
Sumit Taank - San Jose CA, US

Wal-Mart Stores, Inc. - Bentonville AR

G06F 17/30

707728

A computer-implemented method is disclosed for cutting off listings or search results. In the method, a computer system may receive a query from a client computer. In response to the query, the computer system may compile listings ranked according to corresponding relevance scores. Thereafter, the computer system may calculate a cutoff score that is less than a maximum relevance score corresponding to the listings. For example, the computer system may calculate a cutoff score by multiplying a maximum relevance score corresponding to the listings by a weighting. In selected embodiments, the weighting may be greater than 0.05 and less than 0.5. The computer system may then serve to the client computer as a response to the query a subset of the listings. The subset may comprise those listings corresponding to a relevance score greater than the cutoff score.

20140278797, Sep 18, 2014

Mar 15, 2013

13/840580

- Bentonville AR, US
Sumit Taank - San Jose CA, US

Wal-Mart Stores, Inc. - Bentonville AR

G06Q 30/02

705 733

A system is disclosed for determining the popularity of elements within categorized data according to a category to which they pertain based on the popularity of element attributes. The categorized data may be a list of products, where the elements may be products classified by product categories and the attributes may be brand names. Key words may be associated with such brands and/or their products. The key words may be words used when referring to the brands and/or their products with respect to their product category. These key words may be applied to additional data from an external source providing an on-line service, where service user-text is recorded in the additional data. Popularity scores may then be assigned to the products based on indications of references to the brands and/or their products, with respect to their product category, derived from the application of the key words to the additional data.

20140244633, Aug 28, 2014

Feb 28, 2013

13/781580

- Bentonville AR, US
Sumit Taank - San Jose CA, US

Wal-Mart Stores, Inc. - Bentonville AR

G06F 17/30

707723

A computer-implemented method is disclosed for sorting listings or search results. In the method, a computer system may receive a query from a client computer. The computer system may respond to the query with listings ranked according to corresponding relevance scores. Thereafter, the computer system may receive a request to sort the listings according to some numeric attribute (e.g., price) corresponding thereto. Accordingly, the computer system may calculate a modified sorting criterion for each of the listings. Each modified sorting criterion may be weighted by both the original relevance score and the numeric attribute of the corresponding listing. The computer system may then serve, to the client computer, the listings ranked according to the modified sorting criteria.