Highlights on Friday, 8.3.2019
On Friday, a keynote from the industry will open the last day of the conference. Blockchains seem to be a future technology in many business applications that transcend company boundaries. Many applications use blockchain technology to improve processes. The SAP keynote will address the question of what is behind this hype and what technologies are behind this intelligent and secure data distribution. The tasks of traditional databases in this context will also be shown.
The last session of the conference, starting at 10:30 a.m., will have a title that could also be the title of the entire conference and the title of each contribution: "Challenges in Data Processing". The two challenges in this session are generating SPARQL queries from natural language and Partial Reload of Incrementally Updated Tables in Analytic Database Accelerators.
From 11:30 a.m. on, the awards ceremonies of the conference will take place (directly without a break afterwards). First, the dissertation prizes of the GI department DBIS will be awarded. At BTW 2019, for the second time in the history of this award, there will be three winners: The prizes go to Thorsten Papenbrock (Hasso Plattner Institute), Ismail Oukid (TU Dresden, now SAP Walldorf) and Caetano Sauer (TU Kaiserslautern, now Tableau Munich). The award-winning dissertations will be presented by the prizewinners in a short lecture. Furthermore, the best contribution of the conference and the best demonstration will be awarded. The prizes already awarded during the week for the best student contribution and the first three places in the Data Science Challenge will be mentioned once again.
And after the BTW is before the BTW. The last session concludes with a short preview of the BTW 2021, the organising university and working group and the first plans for the next conference.
Highlights on Thursday, 7.3.2019
Thursday opens with the keynote speech of Ihab Ilyas (University of Waterloo) with the topic "Machine-Learning-Procedures for Data Cleaning". Machine learning tools are becoming more and more important, they are also seen as a solution for data curation tasks. The lecture will focus on two topics: Entity consolidation and the use of probabilistic inference to find anomalies and repair data sets.
Session 6 on Query Processing and Optimization will start at 11:00 a.m.. With the title keywords: "Yes-(too-)SQL ..." the first authors are certainly majority capable within our community, ... it will be about a SQL script language for main memory database systems. In the next lecture we will discuss query optimizations through on-the-fly reconfiguration of query plans for the processing of streams. The technology for monitoring load forecasts for intraday electricity trading will follow as an industry contribution to this topic.
Session 7, which also starts at 11.00 a.m., deals with the topic of Similarity. With the increasing interest in machine-learning techniques, there are three contributions in this session that focus on similarity in different applications: the efficient use of Jaro-Winkler-Similarity for search, different similarity measures for Entity Resolution and in the last contribution Nearest Neighbour Joins for relational databases.
The sponsor tutorial on Actian Vector also starts at 11.00 am. The tutorial introduces a database system for analyses on large amounts of data, which is in the tradition of MonetDB and Vectorwise. Actian Vector enables in-the-moment analytics without the limitations of traditional data warehouses. The tutorial introduces the internals of the system. Another sponsor tutorial follows at 3:30 pm: Exasol presents its database solution for scalable analyses on large amounts of data. This tutorial is Hands-On, there is a limited number of seats available in a PC pool, so please register in advance on the website.
In Session 8, Machine Learning, a contribution in the scientific program "Gradient Descent and Tensor Algebra for Main Memory Database Systems", a contribution with authors from science and industry "Efficient Data-Parallel Cumulative Aggregates for Large-Scale Machine Learning" and a contribution from industry will be presented: "Machine Learning Applied to the Clerical Task Management Problem in Master Data Management Systems". Session 9 will then be followed by the topics "Holistic Metadata Management in Data Lake" and the presentation of the DFG priority program "Scalable Data Management for Future Hardware" (SPP 2037), which will be followed by a visit to the poster exhibition of the SPP integrated into the session.
In the afternoon the next 8 performances will take place in the demo program. There will also be a machine learning theme: "Density-based Interactive Clustering and Exploration". As always, the range of software demonstrations will be very broad: from "Big graph analysis", "Connecting a Bruker Mass Spectrometer to the Internet" to "Precision Oncology" and "jHound: for profiling JSON data". The demo program will run on Thursday from 13.30 - 17.00.
At 17:00 the annual general meeting of the FGDB of the Gesellschaft für Informatik takes place. Here the next activities of the department will be presented.
Afterwards we will move to the city centre. At 19.00 the conference dinner will take place in the Radisson Blu.
Highlights on Wednesday, 6.3.2019
The conference will open on Wednesday. After the workshop and tutorial program of the first two days, the scientific and industrial program as well as the presentation of software prototypes will begin today. Between all the topics, there will always be plenty of time for technical discussions with colleagues* in the atrium of the computer science building.
The day will begin with a keynote speech by Stefanie Rinderle-Ma. The focus here will be on processes for integrated data analysis. Data from machines, sensors and information systems will be processed and evaluated by processes, the keynote will establish the connection to the field of Industry 4.0.
At this conference, we decided to bundle the lectures in the individual sessions according to topics, i.e. to combine contributions in the scientific program and in the industrial program. We hope that such interesting considerations of a topic will result from the point of view of science and from the point of view of practice.
The first session is a single-track session under the title "High-Performance Queries" and starts with a contribution from the scientific program and a topic from the industry program. A lecture on the fight against duplicates in hashing will open here. In the second lecture adaptations of the database system Hyper for tableau workloads will be presented.
In the afternoon there will be two parallel sessions on the core topic of the database area "Query Processing and Optimization" and on the topic "Text".
This will be followed by a panel discussion on "Database Systems in the Age of AI and Data Science". Under the direction of Volker Markl, the participants Erhard Rahm, Bernhard Mitschang, Thomas Seidl and Wolfgang Lehner will discuss the positioning of database research on the topics of AI and Data Science.
At the same time, the first eight software demos of this year's BTW will be presented from 1.30 pm to 5 pm. The spectrum of topics here is very broad, ranging today from "Database-Supported Video Game Engines", an "Execution and Learning Tool for Relational Algebra" to the question "When is Harry Potters birthday? And who hasn't always wanted to build a "Self-Brewed Database System"? Today you can exchange ideas with the developers of such a system.
Tonight we stay in the house: At 6.30 pm the get-together starts with a poster session in the atrium of the Konrad-Zuse-Haus. Here we have the opportunity to continue all the technical discussions of the day and to finish off with good food and drinks. Many contributions to the conference will be presented by posters during the Get-Together, so that you can have technical discussions and discussions with the authors of the contributions and get an impression of the topics of the coming days. The evening event is organized by SAP, which supports BTW 2019 as platinum sponsor. SAP will be present in the Atrium and present the SAP Hana Database Campus there.
And finally, a find from a job advertisement published last week: "If you are interested and attend BTW next week: Just get in touch there." If you are interested and attend BTW next week: Just get in touch there." If you are interested and attend BTW next week: Just get in touch there.
Highlights on Tuesday, 5.3.2019
Data Science Challenge
The topic of particulate matter has been on the news for months, and nationwide measurements are available through many sensors in an open science network. At the Data Science Challenge, five teams will compete and present their evaluations of these data; the work will range from determining causal relationships to predicting values. The teams will present their programs and results at 9:00 am.
At 11:00 a.m. Andreas Henrich will give an invited lecture at the Workshop on Big (and Small) Data in Science and Humanities (BigDS) about the state-of-the-art in the field of Digital Humanities. This will be followed by a very interesting workshop program on graph databases, ontologies, entity recognition and text classification.
At 13:30 the tutorial "NoSQL & Real-Time Data Management in Research & Practice" starts. The focus is on the combination of push-based events, as used in stream data systems, with the well-known semantics of queries on traditional databases.
From 11:00 - 17:00 the student program will take place on Tuesday. Students will present the results of their theses here. Often these are the first lectures of the future scientific junior staff in the subject community. The very active working groups in the database sector in Germany have delegated their best students for this purpose.
The first full-day workshop on Novel Data Management Ideas on Heterogeneous (Co-)Processors (NoDMC) will take place on Tuesday. In line with a current DFG priority programme, it deals with the development of database technologies on new hardware platforms. In this workshop there will be two invited lectures by Erich Focht (NEC Deutschland GmbH) on "The high memory bandwidth vector engine NEC Aurora TSUBASA" and Carsten Binnig (TU Darmstadt) on "DPI: The Data Processing Interface for Modern Networks".
The BTW Welcome Event takes place on a ship (as is usual in a port city). It goes from the Rostock city harbour to Warnemünde; past the harbour and the shipyard. The departure time is at 19:00 o'clock, the "boarding" begins already from 18:30 o'clock. Between 21:30 o'clock and 22:00 o'clock we are back at the berth in the city harbour and can end the evening on board comfortably. Attention: this boat tour is fully booked one week before the BTW!
Highlights on Monday, 4.3.2019
BTW 2019 starts on Monday at 13.30 o'clock with 3 parallel programs.
For all those who are already in Rostock we have a tip: a short walk through the city centre and the Rostock city harbour. Most of the hotels are centrally located, close to the university square. In the middle of the pedestrian zone of the city is the main building of the university (Universitätsplatz 1). In the right wing of the main building, there is the treasure chamber of the university; here, there are also the historical matricels, in which since 1419, all enrolled students were recorded in handwriting.
The Marienkirche (address: Bei der Marienkirche 1) is also very worth seeing. If you have a short time to visit the church, you should definitely stop by the astronomical clock. Since 1472, this eleven-metre-high work of art has set the pace for the city of Rostock and shows the time, the day of the week, sunrise and sunset, the phases of the moon, signs of the zodiac and much more. Last year it received a new dial for the next 133 years (so it will display the calendar until 2150).
A walk in the Rostock city harbour is beautiful at any time of the day. The former shipyard area becomes more and more the promenade of the city. With many restaurants, boat harbours and a view to the city silhouette, it always goes along the Warnow. The Federal Garden Show will take place in this area in 2025, the city of Rostock has recently been awarded the contract (the address for the navigation device is: 18055 Rostock, Warnowufer, or Am Strande 2.
From 13.30 o'clock the BTW begins then. In the last months we have put together an interesting and varied programme, on Monday it finally starts: the 18th BTW begins. We are looking forward to the arrival of the participants. In the atrium of the Konrad-Zuse-Haus (Albert-Einstein-Str. 22) the registration for the conference will start at 12:00 noon. On the first day we have three parallel offers:
- Graph database systems with the best-known representative Neo4J are finding more and more applications. In a hands-on tutorial by Lena Wiese you can get to know and try them out. This tutorial is limited in space, you should register under this registration page before.
- At the same time, new concepts for "Digitale Lehre im Fach Datenbanken" will be discussed, here the word workshop is to be taken literally, it is about the exchange of experience and participation.
- And there is a third parallel offer: the combination of database technology and statistical-relational artificial intelligence. These two areas are growing ever closer together in data science applications, a detailed tutorial by Tanya Braun will present these topics.
Our Blog from 25.1. to 18.2.2019
From 25th January to 18th February 2019 we have highlighted some BTW highlights and news here every working day. The order of these highlights was not important.