In the Maze of Data Languages

TL;DR

This paper compares the complexity and expressiveness of automata and logic models for data languages, which involve strings and trees with finite labels and infinite data values, to identify suitable regular models.

Contribution

It provides a detailed analysis and comparison of various automata and logic models for data languages, highlighting their relative strengths and limitations.

Findings

Automata and logic models vary in expressiveness for data languages.

Complexity results differ across models, affecting their practical applicability.

Certain models are identified as better candidates for regular data languages.

Abstract

In data languages the positions of strings and trees carry a label from a finite alphabet and a data value from an infinite alphabet. Extensions of automata and logics over finite alphabets have been defined to recognize data languages, both in the string and tree cases. In this paper we describe and compare the complexity and expressiveness of such models to understand which ones are better candidates as regular models.