My Research

A. IMPROVING SCALABILITY IN RECOMMENDER SYSTEMS

Recommender Systems make use of the “wisdom of the crowd” to help users identify useful items from a large search space. A common technique used by many recommender systems is collaborative filtering., which analyses past community opinions to find correlations between similar users and items to suggest k personalized items to a query user. Today, recommender systems need to handle increasingly large data sets of items and users. To make collaborative filtering scalable, we have explored various strategies to partition the user and item space. Partitioning strategies include Voronoi diagrams, k-d trees, quadtrees, social graph partitioning and partitioning based on context. Once the data is partitioned, the recommendation algorithm is applied independently in the disjoint partitions. As an insurance against degradation of quality of recommendations, our partitioning strategies adapt to the spatial autocorrelation in the underlying space.

J. Das, S. Majumdar, P. Gupta, S. Datta.

Scalable Recommendations using Decomposition Techniques based on Voronoi Diagrams.

Journal of Information Processing and Management, Elsevier, 58(4), 2021.

J. Das, S. Majumdar, P. Gupta

Collaborative Recommendations using Hierarchical Clustering based on K-d Trees and Quadtrees

International Journal of Uncertainty, Fuzziness and Knowledge-Based Systems, World Scientific, 27(4), pp. 637—668, 2019.

A. Srivastava, A. Jain, A. Jayadev, R. Mukherjee, S. Bhargava and P. Gupta

An Experimental Study of Scalability in Cross-Domain Recommendation Systems, Adv. in Intelligent Systems and Computing, Volume 706, pp. 473--481, Springer, 2017.

V. Koshti, NVS Abhilash, Karanjit S. Gill, N. Nair, M. B. Christian, P. Gupta

Online Partitioning of Large Graphs for Improving Scalability in Recommender Systems. Proceedings, International Conference on Computational Intelligence: Theories, Applications and Future directions, ICCI 2017, Springer.

J. Das, S. Dugar, H. Gupta, S. Majumder, P. Gupta

An Adaptive Approach to Collaborative Filtering Using Attribute Autocorrelation Proc., IEEE IEMCON 2015, Vancouver, Canada, October 2015.

J. Das, S. Majumder, D. Dutt, P. Gupta.

Iterative Use of Weighted Voronoi Diagrams to Improve Scalability in Recommender Systems

Proceedings, PAKDD 2015, Ho Chi Minh City, Vietnam, Springer Verlag LNAI Vol. 9077, Part 1, pp. 605-617, May 2015.

S. Datta, J. Das, P. Gupta, S. Majumder.

SCARS: A Scalable Context-Aware Recommendation System. IEEE C3IT 2015, Kolkata, pp. 1—6, Feb 2015.

J. Das, A. K. Aman, P. Gupta, A. Haider, S. Majumder, S. Mitra

Scalable Hierarchical Collaborative Filtering Using BSP Trees.

Springer Verlag Lecture Notes in Electrical Engineering, Vol. 335, pp. 269--278, March 2015.

J. Das, P. Mukherjee, S. Majumder, P. Gupta

Clustering-Based Recommender System Using Principles of Voting Theory.

IEEE International Conference on Contemporary Computing and Informatics, Mysore, 230—235, November 2014.

A. Dalmia, J. Das, P. Gupta, S. Majumder, D. Dutta

Scalable Hierarchical Recommendations Using Spatial Autocorrelation,

Proceedings, International Conference on Big Data Science and Computing, Beijing, China

(BigDataScience 2014. This paper won an award).

J. Das, S. Majumder, and P. Gupta.

Spatially Aware Recommendations Using K-d Trees.

Proceedings, 3rd IEEE International Conference on Computational Intelligence and Information Technology, CIIT 2013, Mumbai, pp. 167—175, Nov 2013.

J. Das, S. Majumder, and P. Gupta.

Voronoi Based Location Aware Collaborative Filtering,

Proceedings, 3rd IEEE National Conference on Emerging Trends and Applications in Computer Science (NCETACS), pp. 179—183, March 2012.

B. COMPUTATIONAL GEOMETRY

This work investigates the design of efficient algorithms and data structures for various query-retrieval problems defined on geometric objects that involve aggregation. In generalized intersection problems, the objects come aggregated in disjoint groups where the grouping is dictated by theunderlying application. The goal is to preprocess these groups into a data structure so that useful questions about the groups, relative to a query object, can be answered efficiently. In range-aggregate query problems, the goal is to preprocess the objects into a data structure such that the result of applying an aggregation function to the objects intersected by a query object, can be reported efficiently.

*A.S. Das, P. Gupta, K. Kothapalli, K. Srinathan.

Reporting and Counting Maximal Points in a Query Orthogonal Rectangle

Journal of Discrete Algorithms (Elsevier), Vol. 30, 78--95, Jan 2015..

* A.S. Das, P. Gupta, K. Kothapalli, K. Srinathan.

On Reporting the L_1 metric Closest Pair in a Query Rectangle

Information Processing Letters, Vol. 114, Issue 5, 256--263, April 2014.

*P. Gupta, R,. Janardan, Y. Kumar, M.Smid,

"Data Structures for Range-Aggregate Extent Queries",

Computational Geometry, Theory and Applications, Elsevier, Vol. 47, Issue 2C, 329--347, Jan 2014.

*B. Sanyal, P. Gupta, S. Majumder.

"Colored Top-K Range-Aggregate Queries",

Information Processing Letters, Volume 113, Issues 19--21, 777--784, Sep-Oct 2013.

*A.S. Das, P. Gupta, and K. Srinathan.

Counting Maximal Points in a Query Rectangle

Proceedings, 7th Workshop on Algorithms and Computation, WALCOM 2013,

Springer Verlag Lecture Notes on Computer Science Vol. 7748, pp. 65—76, February 2013.

*S. Das, P. Gupta, A.K. Kalavagattu, K. Srinathan, K. Kothapalli, J. Agarwal.

Range Aggregate Maximal Points in the Plane,

Proceedings, 6th Workshop on Algorithms and Computation, WALCOM 2012, Dhaka, Bangladesh,

Springer Verlag Lecture Notes on Computer Science, pp. 52—63, February 2012.

*S. Rahul, P.Gupta, K.S. Rajan.

“Data Structures for Range-Aggregation by Categories”,

International Journal of Foundations of Computer Science, Vol. 27, No. 7, 1707—1728, 2011.

*A.S. Das, P. Gupta, K. Srinathan.

"Data Structures for Extension Violations in a Query Range",

Journal of Mathemathical Modeling and Algorithms, Springer, 10(1), 79—107, 2011.

*A.S. Das, P. Gupta, K. Srinathan, K. Kothapalli.

Finding Maximum Density Axes-Parallel Regions in Weighted Point Sets.

Proceedings, 23rd Canadian Conference on Computational Geometry, Toronto, Canada, pp. 129--134, August 2011.

*S. Rahul, P.Gupta, R. Janardan, K.S. Rajan.

Efficient top-K Queries for Orthogonal Ranges.

Proceedings, WALCOM 2011,

Springer Verlag Lecture Notes on Computer Science, Vol. 6552, pp. 110—121, February 2011.

*S. Rahul, H. Bellam, P. Gupta, K.S. Rajan.

Range-Aggregate Structures for Colored Geometric Objects.

Proceedings, 22nd Canadian Conference on Computational Geometry, 48-52, August 2010.

*P. Gupta, R. Janardan and M. Smid.

"Efficient Non-intersection Queries on Aggregated Geometric Data”,

International Journal of Computational Geometry and Applications, 19 (6), 479--506, 2009.