Slava Novgorodov Homepage

Home

My name is Slava Novgorodov, Ph.D. and Computer Science Researcher.

I am Research Scientist at Meta (Facebook).
I obtained my Ph.D. in Computer Science from Tel Aviv University, under supervision of Prof. Tova Milo.

I am also a visiting lecturer at Tel Aviv University, teaching several courses.

I have different hobbies like learning languages, collecting soccer jerseys and traveling around the world.

Recent updates:
GUDE-AI'24, ECNLP'24 and TODS papers were accpeted.

Dr. Slava Novgorodov - Computer Science Researcher
Dr. Slava Novgorodov
Back to Top

Research

Currently I am Research Scientist at Meta, and most of my current projects are connected to the e-commerce. Mainly my focus is text analsysis and NLP research related to the e-commerce applications.

Prior to Meta, I worked at eBay Research as Applied Research Sceintist and before I was in different e-commerce related companies as Software Engineer and Data Scientist and also did research. I strated doing research during my B.Sc. studies in Mathematics and Computer Science when I joined the Data Management Research Group of Tel Aviv University (supervised by Prof. Tova Milo) as an intern. After finishing my B.Sc. in 2011, I continued to M.Sc. in Computer Science (which I finished in February 2014) and Ph.D. in Computer Science (2014-2018), all supervised by Prof. Milo.

During my master's studies I mainly worked on combination of crowdsourcing and web data management. My M.Sc. thesis title was "Asking the Right Questions in Crowd Data Sourcing".

After working with the regular crowd, I started my research about crowdsourcing with experts. During my Ph.D. I worked mainly on "Data Cleaning using Experts" and "Rules Refinement and Sharing between Experts" projects.
The title of my Ph.D. thesis is "Experts in the Loop in Data Management".

Research interests

  • Text Analysis and NLP
  • E-Commerce
  • Crowdsourcing (and Crowdsourcing to experts)
  • (Web) Data Management
  • Back to Top

    Teaching

    Tel Aviv University

    Algorithms & Applications in Social Networks17/18B; 18/19B; 19/20B; 20/21A; 21/22A; 22/23A; 23/24A; 24/25A;
    Intro to Data Science15/16B; 16/17B; 17/18B; 18/19B; 19/20B; 20/21B; 21/22B; 22/23B; 23/24B;
    Workshop on E-Commerce Applications21/22A; 22/23A; 22/23B; 23/24A; 23/24B; 24/25A;
    Data-Centered Crowd Sourcing14/15B; 15/16A; 16/17B; 17/18A; 18/19A; 19/20A; 20/21B; 21/22B;
    Web Data Management14/15B; 15/16A; 16/17B
    Back to Top

    Publications

    List of my publications is also available at DBLP or Google Scholar


    2024


    Cost-Effective LLM Utilization for Machine Learning Tasks over Tabular Data [paper]
    Yael Einy, Tova Milo, Slava Novgorodov
    GUIDE-AI'24 - June 2024, Santiago, Chile

    Learning Reasons for Product Returns on E-Commerce [paper]
    Miriam Farber, Slava Novgorodov, Ido Guy
    ECNLP'24 - May 2024, Turin, Italy

    Automated Category Tree Construction: Hardness Bounds and Algorithms [paper]
    Shay Gershtein, Uri Avron, Ido Guy, Tova Milo, Slava Novgorodov
    ACM Transactions on Database Systems - Just Accepted


    2023


    Efficiently Archiving Photos under Storage Constraints [paper]
    Susan B. Davidson, Shay Gershtein, Tova Milo, Slava Novgorodov, May Shoshan
    EDBT'23 - March 2023, Ioannina, Greece

    Shop by Image: Characterizing Visual Search in E-Commerce [paper]
    Arnon Dagan, Ido Guy, Slava Novgorodov
    Information Retrieval Journal, Vol. 26, Article 2 (March 2023)

    The Tip of the Buyer: Extracting Product Tips from Reviews [paper]
    Sharon Hirsch, Slava Novgorodov, Ido Guy, Alexander Nus
    ACM Transactions on Internet Technology, Vol. 23, Issue 1, Article 4 (February 2023)


    2022


    PHOcus: Efficiently Archiving Photos [paper]
    Susan B. Davidson, Shay Gershtein, Tova Milo, Slava Novgorodov, May Shoshan
    VLDB'22 - September 2022, Sydney, Australia

    Automated Category Tree Construction in E-Commerce [paper]
    Uri Avron, Shay Gershtein, Ido Guy, Tova Milo, Slava Novgorodov
    SIGMOD'22 - June 2022, Philadelphia, Pennsylvania, USA

    Classifier Construction Under Budget Constraints [paper]
    Shay Gershtein, Tova Milo, Slava Novgorodov, Kathy Razmadze
    SIGMOD'22 - June 2022, Philadelphia, Pennsylvania, USA

    On the Hardness of Category Tree Construction [paper]
    Shay Gershtein, Uri Avron, Ido Guy, Tova Milo, Slava Novgorodov
    ICDT'22 - March 2022, Edinburgh, UK

    Disposal by Design [paper]
    Susan B. Davidson, Shay Gershtein, Tova Milo, Slava Novgorodov
    IEEE Data Engineering Bulletin, Volume 45, Number 1 (March 2022)


    2021


    An Image is Worth a Thousand Terms? Analysis of Visual E-Commerce Search [paper]
    Arnon Dagan, Ido Guy, Slava Novgorodov
    SIGIR'21 - July 2021, Montreal, Canada

    ConCaT: Construction of Category Trees from Search Queries in E-Commerce [paper]
    Uri Avron, Shay Gershtein, Ido Guy, Tova Milo, Slava Novgorodov
    ICDE'21 - April 2021, Chania, Greece

    Improving Constrained Search Results By Data Melioration [paper]
    Ido Guy, Tova Milo, Slava Novgorodov, Brit Youngmann
    ICDE'21 - April 2021, Chania, Greece

    Fixing Wikipedia Interlinks Using Revision History Patterns [paper]
    Tova Milo, Slava Novgorodov, Kathy Razmadze
    EDBT'21 - March 2021, Nicosia, Cyprus

    Category Recognition in E-Commerce using Sequence-to-Sequence Hierarchical Classification [paper]
    Idan Hasson, Slava Novgorodov, Gilad Fuchs, Yoni Acriche
    WSDM'21 - March 2021, Jerusalem, Israel

    Generating Tips from Product Reviews [paper]
    Sharon Hirsch, Slava Novgorodov, Ido Guy, Alexander Nus
    WSDM'21 - March 2021, Jerusalem, Israel


    2020


    Descriptions from the Customers: Comparative Analysis of Review-based Product Description Generation Methods [paper]
    Slava Novgorodov, Ido Guy, Guy Elad, Kira Radinsky
    ACM Transactions on Internet Technology, Vol. 20, Issue 4, Article 44 (October 2020)

    CONCIERGE: Improving Constrained Search Results by Data Melioration [paper]
    Ido Guy, Tova Milo, Slava Novgorodov, Brit Youngmann
    VLDB'20 - August 2020, Tokyo, Japan

    MC3: A System for Minimization of Classifier Construction Cost [paper]
    Shay Gershtein, Tova Milo, Gefen Morami, Slava Novgorodov
    SIGMOD'20 - June 2020, Portland, Oregon, USA

    Minimization of Classifier Construction Cost for Search Queries [paper]
    Shay Gershtein, Tova Milo, Gefen Morami, Slava Novgorodov
    SIGMOD'20 - June 2020, Portland, Oregon, USA

    Inventory Reduction via Maximal Coverage in E-Commerce [paper]
    Shay Gershtein, Tova Milo, Slava Novgorodov
    EDBT'20 - March 2020, Copenhagen, Denmark


    2019


    Learning to Generate Personalized Product Descriptions [paper]
    Guy Elad, Ido Guy, Slava Novgorodov, Benny Kimelfeld, Kira Radinsky
    CIKM'19 - November 2019, Beijing, China

    ReducE-Comm: Effective Inventory Reduction System for E-Commerce [paper]
    Shay Gershtein, Tova Milo, Slava Novgorodov
    CIKM'19 - November 2019, Beijing, China

    WiClean: A System for Fixing Wikipedia Interlinks Using Revision History Patterns [paper] [poster]
    Stephan Goldberg, Tova Milo, Slava Novgorodov, Kathy Razmadze
    VLDB'19 - August 2019, Los Angeles, California, USA

    Generating Product Descriptions from User Reviews [paper] [poster]
    Slava Novgorodov, Ido Guy, Guy Elad, Kira Radinsky
    The Web Conference '19 - May 2019, San Francisco, California, USA

    Query Driven Data Labeling with Experts: Why Pay Twice? [paper] [poster]
    Eyal Dushkin, Shay Gershtein, Tova Milo, Slava Novgorodov
    EDBT'19 - March 2019, Lisbon, Portugal


    2018


    GOLDRUSH: Rule Sharing System for Fraud Detection [paper] [poster]
    Ariel Jarovsky, Tova Milo, Slava Novgorodov, Wang-Chiew Tan
    VLDB'18 - August 2018, Rio de Janeiro, Brazil

    Cleaning Data with Constraints and Experts [paper] [technical report]
    Ahmad Assadi, Tova Milo, Slava Novgorodov
    WebDB'18 - June 2018, Houston, Texas, USA

    Rule Sharing for Fraud Detection via Adaptation [paper] [technical report]
    Ariel Jarovsky, Tova Milo, Slava Novgorodov, Wang-Chiew Tan
    ICDE'18 - April 2018, Paris, France

    Interactive Rule Refinement for Fraud Detection [paper]
    Tova Milo, Slava Novgorodov, Wang-Chiew Tan
    EDBT'18 - March 2018, Vienna, Austria


    2017


    DANCE: Data Cleaning with Constraints and Experts [paper] [poster]
    Ahmad Assadi, Tova Milo, Slava Novgorodov
    ICDE'17 - April 2017, San Diego, California, USA


    2016


    Rudolf: Interactive Rule Refinement System for Fraud Detection [paper] [poster]
    Tova Milo, Slava Novgorodov, Wang-Chiew Tan
    VLDB'16 - September 2016, New Dehli, India


    2015


    The Temp Secretary Problem [paper]
    Amos Fiat, Ilia Gorelik, Haim Kaplan, Slava Novgorodov
    ESA'15 - September 2015, Patras, Greece

    QOCO: A Query Oriented Data Cleaning System with Oracles [paper] [poster]
    Moria Bergman, Tova Milo, Slava Novgorodov, Wang-Chiew Tan
    VLDB'15 - September 2015, Kohala Coast, Hawaii, USA

    Query-Oriented Data Cleaning with Oracles [paper]
    Moria Bergman, Tova Milo, Slava Novgorodov, Wang-Chiew Tan
    SIGMOD'15 - June 2015, Melbourne, Australia

    Managing General and Individual Knowledge in Crowd Mining Applications [paper]
    Yael Amsterdamer, Susan B. Davidson, Anna Kukliansky, Tova Milo, Slava Novgorodov, Amit Somech
    CIDR'15 - January 2015, Asilomar, California, USA


    2014


    Ontology Assisted Crowd Mining [paper] [poster]
    Yael Amsterdamer, Susan B. Davidson, Tova Milo, Slava Novgorodov, Amit Somech
    VLDB'14 - September 2014, Hangzhou, China

    OASSIS: Query Driven Crowd Mining [paper]
    Yael Amsterdamer, Susan B. Davidson, Tova Milo, Slava Novgorodov, Amit Somech
    SIGMOD'14 - June 2014, Snowbird, Utah, USA


    2013


    Answering Planning Queries with the Crowd [paper] [technical report]
    Haim Kaplan, Ilia Lotosh, Tova Milo, Slava Novgorodov
    VLDB'13 - August 2013, Riva del Garda, Italy

    CrowdPlanr: Planning Made Easy with Crowd [paper] [poster]
    Ilia Lotosh, Tova Milo, Slava Novgorodov
    ICDE'13 - April 2013, Brisbane, Australia


    2012


    Asking the Right Questions in Crowd Data Sourcing [paper] [poster]
    Rubi Boim, Ohad Greenshpan, Tova Milo, Slava Novgorodov, Neoklis Polyzotis, Wang-Chiew Tan
    ICDE'12 - April 2012, Washington DC, USA


    2011


    Diversification and Refinement in Collaborative Filtering Recommender [paper] [full version]
    Rubi Boim, Tova Milo, Slava Novgorodov
    CIKM'11 - October 2011, Glasgow, Scotland

    DiRec: Diversified Recommendations for Semantic-less Collaborative Filtering [paper] [poster]
    Rubi Boim, Tova Milo, Slava Novgorodov
    ICDE'11 - April 2011, Hannover, Germany

    Back to Top

    About

    Short Bio

    2022/03 - current Research Scientist at Meta
    2018/09 - 2022/03 Applied Research Scientist at eBay Research
    2017/01 - 2018/08 Senior NLP Engineer and Data Scientist at Twiggle
    2014/10 - 2018/10 Ph.D. student in Computer Science, Tel Aviv University
    2014/10 - 2016/12 Chief Scientist at Linicom
    2011/10 - 2014/02 M.Sc. student in Computer Science, Tel Aviv University
    2010/01 - 2014/09 Software Engineer in VisionMap
    2006/10 - 2011/01 B.Sc. student in Mathematics and Computer Science, Tel Aviv University (during the high school)
    2006/10 - 2009/12 Software Developer in IsraDon
    2002/09 - 2008/07 Student in Shevah Mofet High School

    Important dates

    I was born in November 1989, in Russia (Soviet Union). I moved to Israel in September 2001.
    In December 2009 I married my lovely wife Yana, and in October 2011 my first son, Dennis, was born.
    In February 2017 my second and third kids, Alice and Tom, joined the family.

    Industrial experience

    In September 2006, my first job was IsraDon - largest Russian bookstore chain in Israel. My goal was clear - to define, design and implement the IT infrastructure from scratch. I started from buying equipment (computers, printers, barcode scanners, routers & switches), and then moved to the most important part - developing ERP software. I decided to develop a web-based application (mainly because of my passion to the web), using PHP & MySQL as a backend and HTML, CSS, JavaScript and Ajax for the frontend. After finishing developing and deploying the system, my responsibility was managing it and adding a new features. I had a great time there, but in the end of 2009 I decided to move on.

    In January 2010 I started my work in VisionMap - a leading provider of digital mapping systems. I joined the Software Infrastructure Team as a C++ developer, and my responsibilities were developing and deploying the best infrastructure for all software developers of VisionMap. It includes bringing the latest version of C++ libraries and tools to the codebase (boost, Qt, OpenCV, Visual Studio), switching all of our projects to support Unicode (personally, I believe in UTF-8 everywhere), managing and upgrading the continuous integration server, source control, review board and a company's wiki. In addition, I was developing digital mapping applications using our own infrastructure.

    In October 2014 I moved to Linicom - a company that develops platform for online advertising and served as a Chief Scientist. I joined Linicom as a part time Research & Development consultant already in November 2013, and during the time I worked on different projects, until October 2014, when I started to work as Head of R&D, managing development and reasearch proccess in the company. This work included leading of user behaviour research, including better ads targeting and better user experience. Starting January 2016 I moved to the Chief Scientist position and focused on the Data Science research - from prototype to production. We had different research projects, including Information Extraction, Recommendation Systems, Natural Language Processing and Web Data Management in general.

    In January 2017 I joined Twiggle that develops a Natural Language search for e-commerces as a Senior NLP Engineer and Data Scientist. I spent first couple of month in the NLP team, developing features for better search queries understanding, and in parallel learning company's internal proccesses. Later, I moved to the Data Infrastructure team, as the first Data Scientist in the team. I was responsible for researching and developing automatic tools for matching customers data (product catalogs) to Twiggle's Knowledge Graph (Ontology), and fast customers onboarding. In addition, I was working on data-driven tools for semi-automatic knowledge creation, and search results improvements. Finally, I have participated in several innovation and strategic data-related projects.

    In September 2018 I joined eBay Research as Applied Research Scientist were I worked on various NLP and ML project that are aimed to improve understanding of products data, estimating its quality and making eBay's catalog cleaner and better. During my work at eBay Research I also published 10+ research papers related to the e-commerce.

    Starting March 2022 I work at Meta as Research Scientist.

    Hobbies

    In addition to my research and work, I have several hobbies.
    First of them is strongly related to football:
    I started collecting football teams shirts (jerseys) from around the world, each of them having their own story. Later, I decided to focus only on CSKA Moscow (match worn) shirts. I have one of the biggest collections of CSKA Moscow shirts (web-site will be available soon). I also collect football balls and football teams scarfs. And of course, I play and watch football.

    A second hobby of mine is learning languages:
    As most of people that came to Israel from Russia, I have 3 main languages that I speak: Russian, Hebrew and English.
    I spent some time learning Arabic, one of the most spoken languages in the world and also specificly very popular in the Middle East. Currently, I consider it as my fourth strongest language.
    I studied Serbo-Croatian which is a beatiful slavic language, spoken by almost 20 million people, mostly in ex-Yugoslavian republics. There are lots of great music, movies and books in this language.
    I also have some basic knowledge and experience with Spanish. On the slavic languages axis I "played" with Polish and Czech.

    The third one is connected to the previous two - traveling around the world:
    After moving to Israel in 2001, I have visited different countries and places. For some of them I studied a new language, and from part of them I bring football jerseys.
    List of visited places (with dates and photos).

    Back to Top

    Contact

    You can email me for questions related to my research or any other question. My contact information is listed here.

    Back to Top