Navneet Panda
Navneet Panda
Navneet Panda is a software engineer at Google based out of Mountain View, California. He has authored nine patents for search algorithms and other technologies he developed for Google Search. [1]
Career
Navneet Panda has been a software engineer with Google since January, 2007.
At Google he focuses on search quality improvements, low quality demotion (aka Panda demotion), user intent, and localization.
Patents
On August 12, 2013 Navneet Panda and James Kunz filed for Website duration performance based on category durations (patent 9,514,194). The patent's application would be to create website duration performance scores for sites that represent a category-dependent measure of the usefulness of the information available resources found in websites. They filed a patent of the same name on September 24, 2015.
On December 31, 2012 Navneet Panda filed for a Revising search queries (patent 9,449,095).
The application of the patent would allow Google Search two combine terms from two consecutive searches entered by a user on the same topic into a new revised query which would yield more accurate results.
On December 27, 2012 Navneet Panda and April Lehman filed for a patent for Selectively generating alternative queries (patent 9,135,307).
The patent's application would be to allow Google Search to return results from an alternate query if the top ranked results for the first query came from low quality sites.
This application would allow Google to improve the quality of its searches through filtering out results from low quality sties.
On September 28, 2012 Navneet Panda and Vladimir Ofitserov filed for Ranking search results (patent 8,682,892).
The patent was the basis for the Google Panda search algorithm which was used to lower the rank of "low-quality sites" or "thin sites", and return higher-quality sites near the top of the search results. The technology Navneet Panda pioneered allowed the Google to algorithmically assess websites by many of the same quality categories (including site speed and content’s uniqueness and value) initially used by Google’s human website testers. On December 31, 2012 Navneet, Vladmir and Kaihua Zhu filed for another patent of the same name.
On June 27, 2012 Navneet Panda and April Lehman filed for a Site quality score (patent 9,031,929).
The patent's application was to a create a Site quality score determined by comparing the unique queries associated to site to the the queries where the site was selected by users following a search query.
On June 16, 2011 Navneet Panda, Trystan Upstill, Oleksandr Grushetskyy, Andrei Damien, and Aysel Ogsur filed for a Locally significant search queries (patent 9,348,925).
The patent covered an algorithm that would determine whether the location of the user was significant to a search query.
If so, a local search query would be generated using the general search query and a location phrase to represent the user location.
The final set of search results generated would be a combination of the two queries resulting in an increase in search quality.
On March 30, 2007 Navneet Panda, Yi Wu, Jean-Yves Bouget, and Ara Nefian filed a patent for Learning concept templates from web images to query personal image databases.
The algorithm would allow for retrieved images to generate one or templates which could be used to search an image database based on features commonly shared between sub-images of the retrieved images.
List of Publications
Efficient Top-k Hyperplane Query Processing for Multimedia Information Retrieval Navneet Panda and Edward Y. Chang ACM International Conference on Multimedia, MM Oct. 2006
Concept Boundary Detection for Speeding up SVMs Navneet Panda, Edward Y. Chang and Gang Wu International Conference on Machine Learning, 2006
KDX: An Indexer for Support Vector Machines Navneet Panda and Edward Y. Chang (Transactions of Knowledge and Data Engg., Jun 2006)
Exploiting Geometry for Support Vector Machine Indexing Navneet Panda and Edward Y. Chang (2005 SIAM International Conference on Data Mining)
Hypersphere Indexer Navneet Panda, Edward Y. Chang and Arun Qamra Database and Expert Systems Applications, DEXA 2006
Active Learning in Very Large Databases Navneet Panda, Kingshy Goh and Edward Y. Chang (Journal of Multimedia Tools and Applications Special Issue on Computer Vision Meets Databases (invited submission))
Formulating Context-dependent Similarity Gang Wu, Navneet Panda and Edward Y. Chang ACM International Conference on Multimedia (MM), Singapore, November 2005
Formulating Distance Functions via the Kernel Trick Gang Wu, Navneet Panda and Edward Y. Chang ACM International Conference on Data Mining and Knowledge Discovery KDD, 2005
Speeding up Approximate SVM Classification for Data Streams Navneet Panda and Ching-Yung Lin (IBM Technical Report, Aug 2005)
Improving Accuracy of SVMs by Allowing Support Vector Control Navneet Panda, Gang Wu and Edward Y. Chang (UCSB Technical Report, 2004)
Education
In 2006, Navneet Panda graduated from the University of California, Santa Barbara (UCSB) with a PhD in Computer Science. While at UCSB Nanveet worked as a Research Intern at Intel and IBM. He graduated from Indian Institute of Technology, Kharagpur with a Master of Science in Maths and Computing in 2002. [1]