- 13:15
Cinema 3

Will Neural Networks kill the Inverted Index?

With search engines like Solr and Elasticsearch becoming mainstream, limitations become apparent. In this talk, you’ll learn about well-known limitations for the inverted index. Limitations like language detection, searching images, finding answers instead of documents. What fun is there in discussing limitations if you’re not discussing improvements. For each limitation, you’ll learn about a possible solution based on Neural Networks. Meet Neural search. In the end, we’ll see if Neural Search and Inverted indexes are at war, who’s winning, or if they can live together.

About the speakers



Fellow at Luminis /Amsterdam

Jettro is a search relevance engineer who specializes in giving customers the best possible search experience.  Jettro has been working with the elastic stack for years, and he knows his way around Solr as well. Next to search, he is experimenting with Machine Learning. Combining Machine Learning with Search to solve one of the hardest search problems, named ranking, is what keeps him going. Jettro believes in the vision of Luminis that: Knowledge is the only thing that increases by sharing. Therefore he writes blogs, gives training and presents at many meetups and conferences.

Back to program