Search Meetup Munich - November Edition

The Munich Search Meetup is all about search technologies. This is the 32nd meetup, sixth and last one in 2019! We'll be hosted by BMW as part of BMWs tech meetup series as this will be second BMW Tech Night who are also sponsoring drinks and food!

Note: As this will happen on BMW premises, you are required to bring an ID in order to pass through the security gate. In addition, pictures will be taken during the meetup. Please help us to speed up the front gate registration by providing name/birth date in advance if possible. You can register at https://www.eventbrite.de/e/search-meetup-munich-tickets-80857808883

Agenda:

18:30 - 19:00 Gathering, grab snacks & drinks
19:00 - 19:30 Analyzing Cases and Issues: Towards an Intelligent
Knowledge Base for Aftersales - by Bernhard Waltl (Data Scientist working on NLP) from BMW
19:30 - 20:00 Enhancing your security investigation experience with Podim - by Sebastian from PAYBACK
20:00 - 20:15 Break
20:15 - 20:45 Numeric Search for Full Text Fields - by Christoph Goller
20:45 - 21:30 Mingling and hanging out

Abstracts:

Analyzing Cases and Issues: Towards an Intelligent
Knowledge Base for Aftersales

TSARA is an internal database system at BMW Group and assists car dealers, who record issues and problems customers face with their BMW vehicles. In the past, those cases were manually inspected by technical specialists. They provided feedback and proposed solutions based on their expertise and an existing knowledge base. As a data science group at BMW we implement services based on NLP and text mining to support information retrieval for customers and technical specialists. We will discuss text mining, search and information retrieval in an industrial context.

Enhancing your security investigation experience with Podim

Podim provides a Plugin for Kibana to enhance your Kibana experience. Utilizing Podim, you can enrich your Dashboards with additional information from external systems on the fly. It is configurable with an easy and powerful UI.

Numeric Search for Full Text Fields

Numeric Fields have a long tradition in Lucene / Elasticsearch. But what about search for numbers in full text fields. I present a solution that allows range queries and approximate search for numbers in full text fields. For a query “6 inch screen” you can e.g. get matches for a document containing 5.9 inch screen. We are using the old term based trie data structure to implement this. But it is now part of an analyzer and produces terms for full text field.