Jefrey Lijffijt, Dr. Sc. (Tech)

I am a FWO [Pegasus]² Marie Skłodowska-Curie Fellow, working at Ghent University's Internet Technology and Data Science Lab.

Broadly I work on personalised interactive data exploration. We design algorithms and systems to extract knowledge and insight from data. In this context, personalisation is about enabling algorithms to take into account what is most relevant (to focus on) and what you already know (to discount for).

Address:

IDLab, Ghent University, Technologiepark-Zwijnaarde 19, B-9052 Ghent, Belgium

Telephone:

+32 474 539 464

Email:

jefrey.lijffijt@ugent.be

You can find me on Twitter, Facebook, LinkedIn, and Google Scholar.

Recent activity (2018)

Organization

• The European Conference of Machine Learning and Principles and Practices of Knowledge Discovery in Databases (ECML-PKDD) will be hosted in Ghent in September 2020.
• We are organizing another edition of the Workshop on Interactive Data Exploration and Analytics at ACM SIGKDD 2018 (IDEA @ KDD)

Recent and upcoming travels

• Attending the Dagstuhl Seminar on Automating Data Science, Wadern, Germany, September 30 - October 5, 2018.
• Attending the ECML-PKDD Conference, Dublin, Ireland, September 10 - 14, 2018.
• Attending the ACM SIGKDD Conference, London, UK, August 19 - 23, 2018.
• Visiting the Research Unit for Variation, Contacts and Change in English (VARIENG), University of Helsinki, June 4 - 9, 2018.
• Attending the IEEE ICDE Conference, Paris, France, April 16 - 19, 2018.

Publications

• Bo Kang, Jefrey Lijffijt, Raúl Santos-Rodríguez, Tijl De Bie. SICA: Subjectively Interesting Component Analysis. Data Mining and Knowledge Discovery 32(4): 949-987, 2018. (Original [open access])
• Jefrey Lijffijt, Bo Kang, Wouter Duivesteijn, Kai Puolamäki, Emilia Oikarinen, Tijl De Bie. Subjectively Interesting Subgroup Discovery on Real-valued Targets. To appear in Proceedings of the IEEE International Conference on Data Engineering, 2018. (Extended version at arXiv, Poster)
• Kai Puolamäki, Emilia Oikarinen, Bo Kang, Jefrey Lijffijt, Tijl De Bie. Interactive Visual Data Exploration with Subjective Feedback: An Information-Theoretic Approach. To appear in Proceedings of the IEEE International Conference on Data Engineering, 2018. (Extended version at arXiv)

Pre-prints (arXiv)

• Bo Kang, Jefrey Lijffijt, Tijl De Bie. Conditional Network Embeddings. (Manuscript)
• Florian Adriaens, Tijl De Bie, Aristides Gionis, Jefrey Lijffijt, Polina Rozenshtein. From acquaintance to best friend forever: robust and fine-grained inference of social tie strengths. (Manuscript)

Goto:

Brief bio

Research topics

I am interested in theory and practice of statistical modeling and pattern mining in various data. Currently, I am working mostly on a framework for pattern mining based on subjective interestingness in data analysis. As a side project I am active in analysis of natural language corpora. I have also been working on interactive visual data exploration, graph mining and social network analysis, mining web data and analysis of citation networks. More generally, I am interested in mining interesting/surprising patterns in transactional, sequential, relational data, and graphs, as well as in text mining, natural language processing, information theory, and statistical significance testing.

In 2010, together with Kai Puolamaki and Panagiotis Papapetrou, I proposed a more general definition of patterns that we may mine from data. At roughly the same time, Tijl De Bie was independently working on similar concepts and proposed the same definition. Tijl's framework, known as FORSIED, also included a specific approach to quantify the interestingness of such patterns, motivated by Information Theory. My current work is based on this notion. In principle it could be applied to any data exploration setting, but to execute this in practice, many questions remain. For example, in algorithm design/optimization, visualization (of data and results), interaction design/HCI, and scalability. Hence I am interested in all these areas.

Keywords

Exploratory data mining, interactive data analysis, pattern mining, data visualisation, visual analytics, hypothesis testing, statistical significance, maximum entropy modeling, subjective interestingness, graph mining, natural language corpora.

Career path

• ('17-'18) FWO [Pegasus]² Marie Skłodowska-Curie Fellow, embedded in the Internet Technology and Data Science Lab, Ghent Univesity
• ('16) Previously, I was a Postdoc at the Data Science Lab at Ghent University and iMinds, working with Prof. Tijl De Bie on the FORSIED project
• ('14-'15) Research Associate at the University of Bristol, also with Tijl De Bie
• ('14) Postdoctoral Researcher at Aalto University with Prof. Aristides Gionis and Prof. Samuel Kaski
• ('09-'13) I defended my doctoral dissertation at Aalto University in December 2013, my advisor was Prof. Heikki Mannila. I graduated with distinction and I received an award from Aalto University School of Science for the best doctoral disseration of 2013
• ('08-'09) Before my doctoral studies, I have worked as a consultant in predictive analytics at Crystalloids, Amsterdam, and as a research intern at Philips Research, Eindhoven

Awards

• Best paper award Mining and Learning with Graphs Workshop @ KDD, 2018
• For the paper "From acquaintance to best friend forever: robust and fine-grained inference of social tie strengths", Florian Adriaens, Tijl De Bie, Aristides Gionis, Jefrey Lijffijt, Polina Rozenshtein. (Manuscript)
• Participant at the Heidelberg Laureate Forum, 2016
• Best Doctoral Dissertation of 2013, Aalto University School of Science
• Dr. Sc. (Tech.) degree awarded with distinction (highest mark possible) at Aalto University School of Science, Dec 2013
• For my PhD Thesis "Computational methods for comparison and exploration of event sequences". (Original)

Research grants

• Marie Skłodowska-Curie Fellowship (FWO-EU COFUND) Personalised, interactive, and visual exploratory mining of patterns in complex data (2017-2019).
• Co-supervisor FWO Project Data mining without spilling the beans: Preserving more than privacy alone (2017-2020 PI Prof. Tijl De Bie).
• Advisor in Academy of Finland Project Reassessing language change: the challenge of real time (2014-2018, PI Prof. Terttu Nevalainen).
• Funded PhD Position 2012-2013, Finnish Doctoral Programme in Computational Sciences.

Tutorials, panels and invited talks

• Jefrey Lijffijt, Tijl De Bie. Mining Subjectively Interesting Patterns in Data. Tutorial at ECML-PKDD 2018. (Website with slides)
• Jefrey Lijffijt. Personalised Pattern Mining. Invited talk at DECADE Symposium, LIRIS/CNRS, Lyon, 2017. (Slides)
• Jefrey Lijffijt. Word occurrence distributions, and their implications for statistical analysis and comparison of corpora. Invited talk at University of Helsinki, Department of Modern Languages, 2016. (Slides)
• Jefrey Lijffijt, Eirini Spyropoulou, Tijl De Bie. Making Sense of Relational Data. Tutorial at ECML-PKDD 2015. (Website with slides)
• Jefrey Lijffijt. ''So what?'' -- On acing your PhD. Invited talk at the doctoral consortium of ECML-PKDD 2015. (Annotated presentation)
• Stefan Evert, Gerold Schneider, Vaclac Brezina, Stefan The. Gries, Jefrey Lijffijt, Paul Rayson, Sean Wallis, Andrew Hardie. Corpus statistics: key issues and controversies. Panel discussion at Corpus Linguistics 2015, Lancaster, UK. (Slides)
• Jefrey Lijffijt. Are you talking Bernoulli to me? Significance testing and burstiness of words in text corpora. Department of Mathematics and Statistics, University of Jyväskylä, 11 November 2011, Jyväskylä, Finland. (Presentation)

Community services

Organisation of conferences, workshops, panels

• Jefrey Lijffijt, Minsuk Kahng, Remco Chang, Polo Chau, Dafna Shahaf, Christos Faloutsos. KDD 2018 Workshop on Interactive Data Exploration and Analytics (KDD-IDEA). Workshop at ACM SIGKDD 2018, London, UK. (Website)
• Jefrey Lijffijt, Polo Chau, Jilles Vreeken, Matthijs van Leeuwen, Dafna Shahaf, Christos Faloutsos. KDD 2017 Workshop on Interactive Data Exploration and Analytics (KDD-IDEA). Workshop at ACM SIGKDD 2017, Halifax, Canada. (Website)
• Stefan Evert, Gerold Schneider, Vaclac Brezina, Stefan The. Gries, Jefrey Lijffijt, Paul Rayson, Sean Wallis, Andrew Hardie. Corpus statistics: key issues and controversies. Panel discussion at Corpus Linguistics 2015, Lancaster, UK. (Abstract)

Editor for journals

• Data Mining and Knowledge Discovery (DAMI)

Senior conference roles

• Program Co-Chair. European Conference of Machine Learning and Principles and Practices of Knowledge Discovery in Databases (ECML-PKDD), 2020.
• Senior Program Committee. ACM International Conference on Intelligent User Interfaces (IUI), 2019.

Reviewer for journals

• Nature Communications
• Data Mining and Knowledge Discovery (DAMI)
• Machine Learning Journal (MLJ)
• IEEE Transactions on Knowledge and Data Engineering (TKDE)
• ACM Transaction on Knowledge Discovery in Data (TKDD)
• International Journal of Corpus Linguistics (IJCL)
• Entropy

Program committee member for conferences and workshops

• ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD), 2015 - 2016, 2018.
• European Conference of Machine Learning and Principles and Practices of Knowledge Discovery in Databases (ECML-PKDD), 2012 - now.
• The Web Conference (WWW), 2018.
• SIAM International Conference on Data Mining (SDM), 2018.
• IEEE International Conference on Data Science and Advanced Analytics (DSAA), 2018.
• International Joint Conference on Artificial Intelligence (IJCAI), 2016, 2018.
• International Conference on Artificial Intelligence and Statistics (AISTATS), 2018.
• International Symposium on Intelligent Data Analysis (IDA), 2016 - 2017.
• ACM SIGKDD Interactive Data Exploration and Analytics Workshop (KDD-IDEA), 2014 - now.
• Symposium on Visualization in Data Science (VDS) at IEEE VIS, 2018.
• Practical Theories of Data Mining Workshop (PTDM), 2012.

Students

• Alexandru Cristian Mara, PhD Thesis, Jul 2018 - now.
• Ruben Moortgat, MSc (Ir.) Thesis, Jun 2018 - now.
• Xi Chen, PhD Thesis, Aug 2017 - now.
• Junning 'Lemon' Deng, PhD Thesis, Apr 2017 - now.
• Ahmad Mel, PhD Thesis, Mar 2017 - now.
• Florian Adriaens, PhD Thesis, Sep 2016 - now.
• Bo Kang, PhD Thesis, June 2015 - now.
• Samir Tendjaoui, research intern, May - Aug 2016.
• Achille Aknin, research intern, Feb - Jul 2016.
• Matthew Bastow, MSc Thesis (patternVis: an interactive visualisation tool for RMiner), Feb - Sep 2015.
• Karmen Dykstra, research intern, July 2014 - Jan 2015.

Teaching

• Assistent to Probability and Statistics, Ghent University, Spring 2018.
• Assistent to Probability and Statistics, Ghent University, Spring 2017.
• Assistent to Probability and Statistics, Ghent University, Spring 2016.
• Guest Lecture Text Mining & NLP (Part of Introduction to AI), University of Bristol, Spring 2015.
• Assistent to Algorithmic Methods of Data Mining, Aalto University, Fall 2012.
• Assistent to Algorithmic Methods of Data Mining, Aalto University, Fall 2011.
• Assistent to Statistical Significance Testing in Data Mining, Aalto University, Spring 2010.

Refereed publications

Journal articles

• Bo Kang, Jefrey Lijffijt, Raúl Santos-Rodríguez, Tijl De Bie. SICA: Subjectively Interesting Component Analysis. Data Mining and Knowledge Discovery 32(4): 949-987, 2018. (Original [open access])
• Matt McVicar, Benjamin Sach, Cédric Mesnage, Jefrey Lijffijt, Eirini Spyropoulou, Tijl De Bie. SuMoTED: An intuitive edit distance between rooted unordered uniquely-labelled trees. Pattern Recognition Letters 79: 52-59, 2016. (Original [open access])
• Jefrey Lijffijt, Eirini Spyropoulou, Bo Kang, Tijl De Bie. P-N-RMiner: A generic framework for mining interesting structured relational patterns. International Journal of Data Science and Analytics, 1(1): 61-76, 2016. (Original [open access], Code, Addendum)
• Jefrey Lijffijt, Terttu Nevalainen, Tanja Säily, Panagiotis Papapetrou, Kai Puolamäki, Heikki Mannila. Significance testing of word frequencies in corpora. Digital Scholarship in the Humanities, 31(2): 374-397, 2016. (Preprint, Original [free access])
• Jefrey Lijffijt, Panagiotis Papapetrou, Kai Puolamäki. Size matters: Choosing the most informative set of window lengths for mining patterns in event sequences. Data Mining and Knowledge Discovery, 29(6): 1838-1864, 2015. (Preprint, Original, Code)
• Jefrey Lijffijt, Panagiotis Papapetrou, Kai Puolamäki. A statistical significance testing approach to mining the most informative set of patterns. Data Mining and Knowledge Discovery, 28(1): 238-263, 2014. (Original, Itemset mining implementation)

Conference articles

• Jefrey Lijffijt, Bo Kang, Wouter Duivesteijn, Kai Puolamäki, Emilia Oikarinen, Tijl De Bie. Subjectively Interesting Subgroup Discovery on Real-valued Targets. To appear in Proceedings of the IEEE International Conference on Data Engineering, 2018. (Extended version at arXiv, Poster)
• Kai Puolamäki, Emilia Oikarinen, Bo Kang, Jefrey Lijffijt, Tijl De Bie. Interactive visual data exploration with subjective feedback: An information-theoretic approach. To appear in Proceedings of the IEEE International Conference on Data Engineering, 2018. (extended version at arXiv)
• Florian Adriaens, Jefrey Lijffijt, Tijl De Bie. Subjectively Interesting Connecting Trees. In Proceedings of the European Conference of Machine Learning and Principles and Practices of Knowledge Discovery in Databases (ECML-PKDD) - Part II, pp. 53 - 69, 2017. (Preprint, Original, Code)
• Tias Guns, Achille Aknin, Jefrey Lijffijt, Tijl De Bie. Direct mining of subjectively interesting relational patterns. In Proceedings of the IEEE International Conference on Data Mining (ICDM), pp 913 - 918, 2016. (Preprint, Original)
• Kai Puolamäki, Bo Kang, Jefrey Lijffijt, Tijl De Bie. Interactive visual data exploration with subjective feedback. In Proceedings of the European Conference of Machine Learning and Principles and Practices of Knowledge Discovery in Databases (ECML-PKDD) - Part II, pp. 214 - 229, 2016. (Preprint, Original)
• Bo Kang, Kai Puolamäki, Jefrey Lijffijt, Tijl De Bie. A tool for subjective and interactive visual data exploration. In Proceedings of the European Conference of Machine Learning and Principles and Practices of Knowledge Discovery in Databases (ECML-PKDD) - Part III, pp. 3 - 7, 2016. (Preprint, Original)
• Bo Kang, Jefrey Lijffijt, Raúl Santos-Rodríguez, Tijl De Bie. Subjectively interesting component analysis: Data projections that contrast with prior expectations. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD). (Preprint hosted by conference, Original)
• Tijl De Bie, Jefrey Lijffijt, Raúl Santos-Rodríguez, Bo Kang. Informative data projections: A framework and two examples. In Proceedings of the European Symposium on Artificial Neural Networks (ESANN), pp. 635 - 640, 2016. (Preprint, Original, Presentation)
• Jefrey Lijffijt, Eirini Spyropoulou, Bo Kang, Tijl De Bie. P-N-RMiner: A generic framework for mining interesting structured relational patterns. In Proceedings of the IEEE International Conference on Data Science and Advanced Analytics (DSAA), pages 1-10. IEEE Computer Society, Washington, DC, USA, 2015. (Preprint, Original, Presentation)
• Matt McVicar, Cedric Mesnage, Jefrey Lijffijt, Tijl De Bie. Interactively exploring supply and demand in the UK independent music scene. In Proceedings of the European Conference of Machine Learning and Principles and Practices of Knowledge Discovery in Databases (ECML-PKDD) - Part III, pages 289-292. Springer-Verlag, Berlin-Heidelberg, 2015. (Demo, Preprint, Original)
• Karmen Dykstra, Jefrey Lijffijt, Aristides Gionis. Covering the egonet: A crowdsourcing approach to social circle discovery on Twitter. In Proceedings of the 9th International AAAI Conference on Web and Social Media (ICWSM), 2015. (Preprint, Original, Code+Data)
• Matt McVicar, Cédric Mesnage, Jefrey Lijffijt, Eirini Spyropoulou, Tijl De Bie. Supply and demand of independent UK music artists on the web. In Proceedings of the 2015 ACM Conference on Web Science, 2015. (Original [open access])
• Jefrey Lijffijt. A fast and simple method for mining subsequences with surprising event counts. In Proceedings of the European Conference of Machine Learning and Principles and Practices of Knowledge Discovery in Databases (ECML-PKDD 2013) - Part I, pages 385-400. Springer-Verlag, Berlin-Heidelberg, 2013. (Preprint, Original, Presentation)
• Jefrey Lijffijt, Tanja Säily, Terttu Nevalainen. CEECing the baseline: Lexical stability and significant change in a historical corpus. In Outposts of Historical Corpus Linguistics: From the Helsinki Corpus to a Proliferation of Resources (Studies in Variation, Contacts and Change in English 10). Research Unit for Variation, Contacts and Change in English, Helsinki, 2012. (Original [open access])
• Jefrey Lijffijt, Panagiotis Papapetrou, Kai Puolamäki. Size matters: Finding the most informative set of window lengths. In Proceedings of the European Conference of Machine Learning and Principles and Practices of Knowledge Discovery in Databases (ECML-PKDD 2012) - Part II, pages 451-466. Springer-Verlag, Berlin-Heidelberg, 2012. (Preprint, Original, Presentation, Poster)
• Turo Vartiainen, Jefrey Lijffijt. Premodifying -ing participles in the parsed BNC. In Corpus Linguistics and Variation in English: Theory and Description, pages 247-258. Rodopi, Amsterdam, 2012. (Preprint, Original, Presentation)
• Jefrey Lijffijt, Panagiotis Papapetrou, Kai Puolamäki, Heikki Mannila. Analyzing word frequencies in large text corpora using inter-arrival times and bootstrapping. In Proceedings of the European Conference of Machine Learning and Principles and Practices of Knowledge Discovery in Databases (ECML-PKDD 2011) - Part II, pages 341-357. Springer-Verlag, Berlin-Heidelberg, 2011. (Preprint, Original, Presentation, Poster)

Workshop articles

• Kai Puolamäki, Panagiotis Papapetrou, Jefrey Lijffijt. Visually controllable data mining methods. In Proceedings of the 2010 IEEE International Conference on Data Mining Workshops, pages 409-417. IEEE Computer Society, Washington, DC, USA, 2010. (Preprint, Original)
• Jefrey Lijffijt, Panagiotis Papapetrou, Jaakko Hollmén, Vassilis Athitsos. Benchmarking dynamic time warping for music retrieval. In Proceedings of the 3rd International Conference on Pervasive Technologies Related to Assistive Environments (PETRA), article 59. ACM New York, NY, USA, 2010. (Preprint, Original, Data, Code, Results)
• Jefrey Lijffijt, Panagiotis Papapetrou, Jaakko Hollmén. Tracking your steps on the track: Body sensor recordings of a controlled walking experiment. In Proceedings of the 3rd International Conference on Pervasive Technologies Related to Assistive Environments (PETRA), article 58. ACM New York, NY, USA, 2010. (Preprint, Original, Data)

Non-refereed publications

Letters to journals

• Jefrey Lijffijt, Stefan Th. Gries. Correction to Stefan Th. Gries' "Dispersions and adjusted frequencies in corpora". International Journal of Corpus Linguistics, 17 (1), 147-149, 2012. (Preprint, Original)

Technical reports

• Jefrey Lijffijt, Panagiotis Papapetrou, Niko Vuokko, Kai Puolamäki. The smallest set of constraints that explains the data: a randomization approach. TKK-ICS-R31, TKK Reports in Information and Computer Science, Espoo, May 2010. (Original)
• Jefrey Lijffijt, Ingrid C. M. Flinsenberg. Compression-based activity classification and motif discovery in time series of acceleration data. TN-2008-00521, Koninklijke Philips Electronics N.V., Eindhoven, September 2008.

Doctoral thesis

• Jefrey Lijffijt. Computational methods for comparison and exploration of event sequences. Doctoral dissertation, Aalto University School of Science, Dec. 2013. (Original)

Master's thesis

• Jefrey Lijffijt. Compression-based activity classification and motif discovery in time series of acceleration data. Master's thesis, Utrecht University, Sep. 2008. (Original)