A dichotomy result for Ramsey quantifiers
| Authors | |
|---|---|
| Publication date | 2015 |
| Host editors |
|
| Book title | Logic, Language, Information, and Computation |
| Book subtitle | 22nd International Workshop, WoLLIC 2015, Bloomington, IN, USA, July 20-23, 2015 : proceedings |
| ISBN |
|
| ISBN (electronic) |
|
| Series | Lecture Notes in Computer Science |
| Event | 22nd International Workshop Logic, Language, Information, and Computation |
| Pages (from-to) | 69-80 |
| Publisher | Heidelberg: Springer |
| Organisations |
|
| Abstract |
Ramsey quantifiers are a natural object of study not only for logic and computer science, but also for formal semantics of natural language. Restricting attention to finite models leads to the natural question whether all Ramsey quantifiers are either polynomial-time computable or NP-hard, and whether we can give a natural characterization of the polynomial-time computable quantifiers. In this paper, we first show that there exist intermediate Ramsey quantifiers and then we prove a dichotomy result for a large and natural class of Ramsey quantifiers, based on a reasonable and widely-believed complexity assumption. We show that the polynomial-time computable quantifiers in this class are exactly the constant-log-bounded Ramsey quantifiers.
|
| Document type | Conference contribution |
| Language | English |
| Published at | https://doi.org/10.1007/978-3-662-47709-0_6 |
| Permalink to this page | |
