Re-examining recipe search
summary
Uses JavaScript and neo4j graph database to index recipes harvested from APIs (BigOven, Edamam, Food2Fork), building a command‑line import and query pipeline that locks down ingredient→recipe relationships. Prototype surfaces good pantry‑to‑recipe matches but APIs only provide on the order of hundreds of thousands of recipes and inconsistent cuisine tags, so cross‑cultural discovery (e.g., West African links to Southern food) remains underpowered. Roadmap: open source the repo, add a web UI, and switch to scraping; the proof‑of‑concept works but is data‑limited.
AI-written summary · about
chapters (12)
- 0:00Introduction
- 0:15Recipe management challenges
- 2:16Searching for the right method
- 4:25Data sources and APIs
- 6:48The value of user comments
- 8:20Culinary connections and history
- 10:04Introduction to graph databases
- 13:42Building the graph model
- 17:24Data collection struggles
- 20:32Project implementation goals
- 23:32Embracing failure as learning
- 25:48Conclusion and thanks
related talks
30mScaling Grandma’s Hefezopf
34mBuild an AI Recipe App: AWS Amplify Gen2 & Amazon Bedrock with Erik H. (Pt 2)
18mBuild an AI Recipe App: AWS Amplify Gen2 & Amazon Bedrock with Erik H. (Pt 1)
38mRecommendation Engine with Neo4j, Piece of Cake
20mJavascript AI Cookbook: Building AI powered product features
23mRevolutionizing API Search with Holon Concept Explorations
38mUnlock the Power of JavaScript’s Hidden Gems
45mHow to cook with RxJS