Xianyu Technology

Dec 12, 2018 · Big Data

Community Data Normalization Using Prefix Matching and Text Similarity

The study presents a four‑step pipeline that normalizes community data for rental platforms by clustering records using longest‑common‑prefix patterns, geographic filtering, Levenshtein similarity, and pattern‑based parent‑child assignment, achieving under 8 % false positives and 5 % false negatives.