VectorMaton: Efficient Vector Search with Pattern Constraints via an Enhanced Suffix Automaton
Haoxuan Xie, Siqiang Luo · 2026
Read VectorMaton: Efficient Vector Search with Pattern Constraints via an Enhanced Suffix Automaton by Haoxuan Xie, Siqiang Luo (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-01525.txt.
Description
Haoxuan Xie and Siqiang Luo tackle search in vector databases when each item also carries a sequence, such as text or DNA. Their 2026 work adds pattern rules, like the database commands that check if a string contains a given piece. Old tools filter by plain traits or ranges, but not by such patterns inside sequences.
Their index, VectorMaton, merges pattern checks with nearest-vector search in one enhanced automaton structure. The index stays close in size to the raw data set, which keeps storage costs sane. Tests on real data sets show far higher query speed than rival tools at the same accuracy. Index size also drops sharply in the tests. The work brings a classic database skill to modern AI search stacks.
Catalog record
| CREATOR | Haoxuan Xie, Siqiang Luo |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 79108 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.01525