TDWI Rountable Munich – a Business Big Data Meetup

On Thursday, 28th of November 2013, I joined the TDWI Roundtable Munich organised by TDWI Europe and Henrik Behrens from the SHS Viveon AG. Thanks a lot for organising this free community event!

Being an active meetup organiser myself, it is great to see other user groups are going well, too. The most interesting aspect of the evening were the differences between the TDWI roundtable Munich and the regularly very technical and application-oriented meetups organised by comSysto.

• About 80 people visited the TDWI roundtable Munich, which are more people than we usually have at our meetups.
• About 15% of the participant were women, which is a much higher rate than we have at our meetups.
• Most people were wearing suits, which is a big difference to our meetup participants.
• The language of all presentations was German. Almost all presentations and communication at our meetups is in English.
• The number of questions and the amount of discussion at the end of every talk was very low. In our meetups, we often have long running discussions and a lot of experience exchange.
• Except for one talk, the talks included a good deal of marketing for products or services. We try to convince speakers at our meetups to skip the marketing part.

As you can see, there are many differences between the TDWI roundtable and the comSysto organised meetups. I guess that in most cases that’s due to the different interests of the user groups. The TDWI user group is very strongly business oriented, while groups organised by comSysto tend to be much more technical.

But nevertheless, it was a great event and provided interesting insides for me. Hendrik Behrens from SHS Viveon AG talked about web crawling and moving unstructured HTML / XML data into structured SQL, and did a nice live demo on Cloudera Impala. Afterwards, Christoph Schmitz from United Internet AG provided some nice details about their text mining and log file parsing algorithms. My favourite part was his argument that many parts of Big Data Analyses are not inherent to Big Data. For instance, he and his colleagues spend most of the time training models on small datasets on their laptops. Only for aggregating data and for predictions are Big Data technologies such as Hadoop required. The last two talks were from IBM and SaS, presenting their products and solutions for text mining with a huge amount of data.

I think it’s great to see that classical Data Warehousing groups such as TDWI are opening up to technologies like Hadoop. I enjoyed the evening, did some nice networking and I hope that there will be more TDWI roundtables in Munich in 2014. We might even set up some cooperation to spread the word about each other’s user groups.