#Lucene now uses SIMD instructions to decode postings lists (at last!) issues.apache.org/jira/browse/LU…

We just released GRF version 1.0.0 to CRAN, which officially brings the package out of 'beta'. Highlights include support for 'best linear projection', as well as significant performance improvements. Full list of changes: bit.ly/grf-100

Wherein, I recant my heretical beliefs about the neural hype: sigir.org/wp-content/upl…

From the "ivory tower" to the "real-world": the story of how block-max WAND made its way into Lucene 8 by Adrien Grand et al. is told in a #ecir2020 preprint: cs.uwaterloo.ca/~jimmylin/publ… Lessons for academics seeking to achieve research impact?

Learning more about semantic search started to get interesting when I realized that a widely used search dataset should not be used to evaluate semantic search algorithms due to high term-matching bias. #nlp #NLProc #searchengine towardsdatascience.com/why-you-should…

We’re celebrating the power of open source with a historical timeline of how #Lucene became the search engine that powers digital experiences worldwide. #20YearsOfLucene go.es.io/2VE5iUX

#Elasticsearch 7.7 is coming, and with it come dramatic heap usage improvements. This means reduced storage costs for large data sets and increased cluster stability. Learn about this exciting update. go.es.io/2yAFAJb

I'll be asking "Is it possible for language models to achieve language understanding?" at an upcoming Stanford HAI/OpenAI workshop on GPT-3. My current answer: "We don’t currently have compelling reasons to think they can't": medium.com/@ChrisGPotts/i…

BREAKING: In major copyright battle between tech giants, SCOTUS sides w/ Google over Oracle, finding that Google didnt commit copyright infringement when it reused lines of code in its Android operating system. The code came from Oracle's JAVA SE platform. supremecourt.gov/opinions/20pdf…

GRF 2.0.0 is now available on CRAN. This release adds support for survival analysis, multi-arm causal forests, and more. Full list of changes: grf-labs.github.io/grf/CHANGELOG.…. We welcome your feedback on GitHub!

Elasticsearch and Lucene often take inspiration from academic work. This post shares strategies for implementing research papers, drawing on our experiences: elastic.co/blog/implement…

Lucene 9.0.0 is out! lucene.apache.org/core/corenews.…

Forthcoming #wsdm2022 demo where Josh Devins Julie Tibshirani and I share thoughts about aligning the research and practice of building search applications. Within the Lucene ecosystem, it is possible to bridge the two worlds - to have your cake and eat it too! cs.uwaterloo.ca/~jimmylin/publ…

Elasticsearch 8.0 improves the scalability of vector search with fast approximate nearest neighbor (ANN) search. Read our blog to learn more about this capability: go.es.io/3Lgyudk

GRF 2.1 is now available on CRAN. This release introduces the RATE metric for validating treatment heterogeneity, and extends causal survival forests so they can be used to target survival probabilities. Let us know what you think on GitHub! github.com/grf-labs/grf

Do LLMs understand concurrency? With good code context they kinda do! Julie Tibshirani surfaced this strange comment suggestion from Cody. Turns out with the context of the broader code, Cody was able to infer that the line in question did indeed introduce a race condition!

When running data-intensive services like code search engines, it's critical to monitor container memory. But what cAdvisor memory metrics should you use -- total usage, working set, or resident set size? sourcegraph.com/blog/beyond-wo…

Awesome blog post by Julie Tibshirani! I loved this section, it’s rare to see so well written technical posts these days. sourcegraph.com/blog/keeping-i…