SISAP 2020: 13th International Conference on Similarity Search and Applications

Authors: Rasmus Pagh 

Website: http://www.sisap.org

Dates: September 30 - October 2, 2020

Location:On-line

General Chair: Martin Aumüller, Björn Þór Jónsson, and Rasmus Pagh
Program Committee Co-Chairs: Shin'ichi Satoh, Lucia Vadicamo, Arthur Zimek
Doctoral Symposium Chair: Ilaria Bartolini
Publication Chair: Fabio Carrara
Local Arrangements Chair: Julie Tollund


The 2020 edition of SISAP was planned to be held at IT University of Copenhagen, Denmark, but was converted into an online event due to the on-going pandemic.

A strong technical program was assembled by three program committee co-chairs, 63 program committee members, and 18 additional reviewers. Each of 50 valid submissions, with authors from 22 countries, was reviewed by at least three referees. 31 papers were accepted, 12 of them as short papers. The doctoral symposium accepted 2 papers. 

Gallery view from the special session on Artificial Intelligence and Similarity

The program included four regular sessions, the doctoral symposium, and a special session on Artificial Intelligence and Similarity, chaired by Anshumali Shrivastava, with four talks followed by a panel discussion. The technical program was completed with three distinguished keynote speakers:

  • Marcel Worring from the University of Amsterdam spoke about Interactive Exploration using Hypergraphs. In his engaging presentation, Marcel focused on an interactive exploration of large multimedia collections. He first reviewed recent successes in supporting scalable categorisation, and then highlighted the opportunities provided by the new field of hypergraph learning.
  • Divesh Srivastava from AT&T Labs-Research spoke about Exploiting Similarity Relationships to Repair Graphs. In an entertaining talk, Divesh showed how similarity concepts are important in data management tasks such as entity resolution and taxonomies for noisy data.
  • Ilya Razenshteyn from Microsoft Research spoke about Scalable Nearest Neighbor Search for Optimal Transport. The Wasserstein (aka Optimal Transport) distance is a popular similarity measure for structured data domains, modelled as collections of point sets. The talk focused on efficient algorithms for approximating the distance between a pair of point sets, showing both theoretically well-founded and practical results.

The program committee identified five papers as candidates for the best paper award. It was decided to give the award to Vladimir Mic and Pavel Zezula for their paper “Accelerating Metric Filtering by Improving Bounds on Estimated Distances”. The best student paper award was given Erik Thordsen and Erich Schubert for the paper “ABID: Angle Based Intrinsic Dimensionality”. The best doctoral symposium paper award was given to Shima Moghtasedi for the paper “Temporal Similarity of Trajectories in Graphs”. Top papers from the conference were invited for a special issue of the journal Information Systems.

116 participants signed up for the conference, about half of them from Europe and the other half from institutions around the world. Due to generous sponsorships from Springer, Google, and the IT University of Copenhagen, we were able to make registration completely free. To allow participation from many time zones, a condensed schedule was used with a 5-6 hour main time slot each day. Speakers provided pre-recorded long versions of their talks and gave a short, interactive version on Zoom during the conference. Most participants were active, with 30-40 participants on average in poster sessions, and 30-60 in the technical sessions.

To facilitate interaction, there were three poster sessions placed such that it was possible to attend two at reasonable hours in any time zone. There was also a social event, featuring a popular quiz about Copenhagen. For the poster and social events, we used the gather.town platform, in which a small virtual conference venue had been built.

The conference venue in gather.town: poster room
The conference venue in gather.town: room for gatherings
A scene from the Copenhagen quiz during the social event.

Acknowledgements: Many people worked hard to make SISAP 2020 a success, despite the challenging circumstances. We are particularly indebted to the PC chairs Shin’ichi Satoh, Lucia Vadicamo, and Arthur Zimek, the doctoral symposium chair Ilaria Bartolini, the publication chair Fabio Carrara, and our local arrangements chair Julie Tollund.

Towards SISAP 2021:

As is traditional, the venue for SISAP 2021 was unveiled during the social event. SISAP 2021 is planned to be held in Dortmund, Germany, with Erich Schubert as general chair. We hope that by fall of 2021, the pandemic has subsided sufficiently to allow us to travel to Dortmund, but the experience from SISAP 2020 should provide a template for an online event. On behalf of the organisers, we thank all authors and participants for their contributions, and look forward to seeing you all at SISAP 2021!

About SISAP:

The International Conference on Similarity Search and Applications (SISAP) is an annual forum for researchers and application developers in the area of similarity data management. It aims at the technological problems shared by numerous application domains, such as data mining, information retrieval, multimedia, computer vision, pattern recognition, computational biology, geography, biometrics, machine learning, and many others that make use of similarity search as a necessary supporting service.

From its roots as a regional workshop in metric indexing, SISAP has expanded to become the only international conference entirely devoted to the issues surrounding the theory, design, analysis, practice, and application of content-based and feature-based similarity search. The SISAP initiative has also created a repository serving the similarity search community, for the exchange of examples of real-world applications, the source code for similarity indexes, and experimental testbeds and benchmark data sets (http://www.sisap.org). The proceedings of SISAP are published by Springer as a volume in the Lecture Notes in Computer Science (LNCS) series.

Bookmark the permalink.