Machine Learning and Information Management Lab

Research group

Machine Learning and Information Management Lab

There are many different aspects to BIG DATA whether challenging Veracity or Variety they all require sophisticated statistical analysis, provided by machine learning. While Volume and Velocity are impossible without efficient technical solutions. Our research interests are spread among these directions. Most of our experience comes from the field of Information Retrieval and Databases.

Main research projects

Theoretical Machine Learning (ML)

Tree models, sequence analysis, ensembles, and GPU enabled ML

Search Engines (SE) and Information Retrieval (IR)

ML for Ranking, SE User Behaviour Analysis, SE performance, SE evaluation and Storage and processing of scientific and graph data

Information Management

Stream processing, declarative computation in BD environment. Efficient storage and index structures, e.g column-oriented DB. Optimization and execution of declarative queries and workflows. The holistic application, optimization, and tuning. Data quality. Consistency and high reliability.

Beside the research projects, we deliver special courses:

Machine Learning @Computer Science Centre (https://compscicenter.ru/courses/machine-learning-1/2016-autumn/, https://compscicenter.ru/courses/machine-learning-2/2017-spring/)
Search Engine Architecture @St.Petersburg State University
Information Management @St.Petersburg State University
Database algorithms @St.Petersburg State University

Students interested in research problems in the areas of our interest are welcome to join our lab. The best way to learn more about our research is to take our courses or attend our open seminars. New projects are launched regularly, sometimes it is also possible to join an ongoing project or extend its scope. Please contact the project leader for information on a specific project.

All students willing to join our projects must be skilled in either statistics or programming, preferably in both. The successful candidates will be invited to join one of the projects as a regular team member.

News

Public speaking master class 4 June 2019
PolyTree framework for tree ensemble analysis 27 May 2019
Aggregation of pairwise comparisons with reduction of biases 20 May 2019

Artem Trofimov, Mikhail Shavkunov, Sergey Reznick, Nikita Sokolov, Mikhail Yutman, Igor E. Kuralenok, and Boris Novikov

Reproducible and Reliable Distributed Classification of Text Streams

In Proceedings of the 13th ACM International Conference on Distributed and Event-based Systems (DEBS '19). ACM, New York, NY, USA, 264-265.
June 2019
Igor Kuralenok, Natalia Starikova, Aleksandr Khvorov, and Julian Serdyuk

Construction of Efficient V-Gram Dictionary for Sequential Data Analysis

The 27th ACM International Conference on Information and Knowledge Management (CIKM ’18), October 22–26, 2018, Torino, Italy. ACM, New York, NY, USA, 10 pages
October 2018
Igor Kuralenok, Artem Trofimov, Nikita Marshalkin, Boris Novikov

Deterministic Model for Distributed Speculative Stream Processing

In: Benczúr A., Thalheim B., Horváth T. (eds) Advances in Databases and Information Systems. ADBIS 2018. Lecture Notes in Computer Science, vol 11019. Springer, Cham, September 2018
Artem Trofimov

Consistency Maintenance in Distributed Analytical Stream Processing

In: Benczúr A. et al. (eds) New Trends in Databases and Information Systems. ADBIS 2018. Communications in Computer and Information Science, vol 909. Springer, Cham, September 2018
Igor Kuralenok, Artem Trofimov, Nikita Marshalkin, Boris Novikov

FlameStream: Model and Runtime for Distributed Stream Processing

BeyondMR'18 Proceedings of the 5th ACM SIGMOD Workshop on Algorithms and Systems for MapReduce and Beyond, June 2018
Igor Kuralenok, Nikita Marshalkin, Artem Trofimov, Boris Novikov

An optimistic approach to handle out-of-order events within analytical stream processing

Proceedings of the Third Conference on Software Engineering and Information Management. Saint Petersburg, Russia. CEUR Workshop Proceedings, 2135, April 2018
G. Chernishev, M. Akhin, B. Novikov, and V. Itsykson

Message from the editors

CEUR Workshop Proceedings, 1864, 2017
Nikita Bobrov, George Chernishev, and Boris Novikov

Workload-independent data-driven vertical partitioning

In Marite Kirikova, Kjetil Nørvåg, George A. Papadopoulos, Johann Gamper, Robert Wrembel, Jérôme Darmont, and Stefano Rizzi, editors, New Trends in Databases and Information Systems - ADBIS 2017 Short Papers and Workshops, AMSD, BigNovelTI, DAS, SW4CH, DC, Nicosia, Cyprus, September 24-27, 2017, Proceedings, volume 767 of Communications in Computer and Information Science, pages 275–284. Springer, 2017
Nikita Bobrov, George Chernishev, Dmitry Grigoriev, and Boris Novikov

An evaluation of TANE algorithm for functional dependency detection

In Yassine Ouhammou, Mirjana Ivanovic, Alberto Abelló, and Ladjel Bellatreche, editors, Model and Data Engineering - 7th International Conference, MEDI 2017, Barcelona, Spain, October 4-6, 2017, Proceedings, volume 10563 of Lecture Notes in Computer Science, pages 208–222. Springer, 2017
G. Shalygina and B. Novikov

Implementing common table expressions for MariaDB

CEUR Workshop Proceedings, 1864, 2017
Vyacheslav Galaktionov, George Chernishev, Boris Novikov, and Dmitry Grigoriev

Matrix Clustering Algorithms for Vertical Partitioning Problem: an Initial Performance Study

Selected Papers of the XVIII International Conference on Data Analytics and Management in Data Intensive Domains (DAMDID/RCDL 2016), Ershovo, Moscow Region, Russia, October 2016