A Shiny application for job matching
summary
Prototype Shiny app matches a user profile to job ads using text mining on 150 PDFs of formatted job descriptions; pipeline uses tokenization, n‑grams, stemming, TF‑IDF then clustering with LDA + k‑means to create a K‑dimensional topic space. Interactive flow moves a user from a central vector through four modules (tag likes/dislikes, five pairwise choices, weighting, validation), narrows 100 → 10 → 6 candidates and returns 6 matched job descriptions; backend is a SQLite backend with user action tracking.
AI-written summary · about
chapters (7)
- 0:00Introduction to job matching app
- 1:55Background and project context
- 7:06Demo of the Shiny application
- 11:16Text mining and algorithm design
- 23:13Shiny modules and architecture
- 34:17User tracking and database
- 36:31Conclusion and Q&A
related talks
20mRpanda trading simulation - from an idea to a multi-user shiny app
20mUsing R, the Tidyverse, H2O, and Shiny to reduce employee attrition
20mShiny in Action: Transforming Film Production with TARS
34mBuilding Data Driven Products with Ruby
53mReal-Time Big Data Analytics with R
20mPython In The Job Market
13mMobile Education App during COVID | Teaching Data Science
37mDecoding Talents - Companies in Transition