Lucene/Solr Revolution 2017

Search is a critical component of any effective website or application, connecting users to the data they need to make decisions, whether it’s to find documents, do research or complete an online purchase. Modern search engines have evolved significantly in the past 5 years, incorporating machine learning and artificial intelligence techniques in all aspects of document and query processing, as well as in user profiling and personalization. These techniques require large volumes of high quality training and evaluation data in order to be effective.

In this session, you’ll hear Kevin Vondemkamp, Vice President – Web, Social & eCommerce of Appen and Grant Ingersoll, CTO of Lucidworks discuss how Lucidworks leverages Appen’s crowdsourcing capabilities to enhance its machine learning technology to improve all aspects of the search experience.

Is "machine learning + open source search + social insects" a bit of a gamble for a talk topic? Absolutely, this conference is in Las Vegas after all!

Join Lucene/Solr committer and beekeeper Christine for an easy step-by-step walk through Apache Solr's Learning-to-Rank plug-in. No prior Solr or machine learning experience needed. We will index some bees, err, I mean honey and bee related tweets, and then use those documents to come up with different machine learnt re-ranking models to improve search results' relevance.

In the past, Sandia National Labs has used several COTS products for enterprise search. A few years ago the decision was made to abandon all COTS Enterprise Search products and use Solr and custom in-house developed applications for our enterprise search needs. The enterprise search marketplace has changed since that decision was made and the expectations of our search users have grown dramatically, making it hard for us to keep up with expectations. Lucidworks Fusion looked to us like a viable product that could be used to help us achieve our enterprise search goals at a reasonable cost. We conducted a three month "Proof of Concept" study to evaluate Lucidworks Fusion against our current and projected future requirements. This talk will discuss how we evaluated Lucidworks Fusion, what we looked at in terms of our requirements, and the results of our evaluation.

FindLectures.com is a discovery engine for tech talks, historic speeches, and academic lectures. The site rates audio and video content for quality, showing different recommended talks each day on a variety of topics.

FindLectures.com crawls conference sites to get talk metadata, such as speaker names and bios, descriptions, and the date a video was recorded. Often these attributes are sparsely populated, or available across multiple websites. Additional attributes are inferred from audio and video content, but require more sophisticated data extraction to be useful in a text- oriented search engine like Solr.

This talk will discuss interesting lessons learned from crawling historical videos, demonstrate information extraction with machine learning, and show how to map real world problems to search engine functionality.