Elasticsearch: you know, for more than search
summary
Elasticsearch and the Elastic stack get used for three concrete pipelines: a graph API that computes statistically significant co‑occurrence (used on MovieLens to recommend Requiem for a Dream over generic hits), a forensic graph mode that traces Conficker to a first admin login, and reactive automation via saved queries / percolation plus Watcher webhooks (a sentiment microservice that auto‑retweets some positive tweets actually fires). Built‑in Machine Learning scores anomalies (0–100 severity) on time‑series bytes and flags multivariate service spikes for alerting.
AI-written summary · about
chapters (11)
- 0:00Intro
- 1:05The Elastic Stack
- 7:40Co-occurrence
- 8:55Graphing Relationships
- 15:48Frequency distributions
- 16:57How Graph Databases typically work
- 18:19How Elasticsearch Graph API works
- 31:43Graph Recap
- 34:41Percolate use cases
- 45:03Anomaly detection
- 48:51Types of anomalies
related talks
35mElasticsearch - Beyond Full-text Search
16mWhy we built ElasticSearch
37mBeyond the basics with Elasticsearch
26mFrom Spark to Elasticsearch and Back
52mGrowing to full-suite observability with Elasticsearch & the ELK Stack (PRT267)
30mDebugging at Scale Using Elastic and Machine Learning
36mRecommendations at scale with Elastic Search
47mMake sense of your (big) data using Elasticsearch