Around Dot Depth Two
نویسندگان
چکیده
It is known that the languages definable by formulae of the logics FO2[<,S], ∆2[<,S], LTL[F,P,X,Y] are exactly the variety DA∗D. Automata for this class are not known, nor is its precise placement within the dot-depth hierarchy of starfree languages. It is easy to argue that ∆2[<,S] is included in ∆3[<]; in this paper we show that it is incomparable with B(Σ2)[<], the boolean combination of Σ2[<] formulae. Using ideas from Straubing’s “delay theorem”, we extend our earlier work [LPS08] to propose partially-ordered two-way deterministic finite automata with look-around (po2dla) and a new interval temporal logic called LITL and show that they also characterize the variety DA∗D. We give effective reductions from LITL to equivalent po2dla and from po2dla to equivalent FO2[<,S]. The po2dla automata admit efficient operations of boolean closure and the language non-emptiness of po2dla is NP-complete. Using this, we show that satisfiability of LITL remains NP-complete assuming a fixed look-around length. (Recall that for LTL[F,X], it is PSPACE-hard.) A rich set of correspondences has been worked out between diverse mechanisms for defining the first-order definable word languages and their subclasses (a recent survey is [DGK08]). In the following, CFA refers to counter-free automata, SFRE to star-free regular expressions and Ap refers to the variety of aperiodic monoids [Pin86]. CFA ≡ SFRE ≡ Ap ≡ FO[<] ≡ LTL[U,S] ≡ IT L Further, Thomas showed [Tho82] that by restricting the quantifier-alternation depth in the FO[<] formulae a strict dot-depth hierarchy of star-free languages is obtained, see the paper by Pin and Weil [PW97] for details. For example, B(Σ2)[<] is the class of languages defined by the boolean combination of Σ2[<] formulae, which are the ones which have one block of existential quantifiers followed by one block of universal quantifiers followed by a quantifierless formula. For the FO formulations below, given an alphabet A and a ∈ A, the unary predicate Qa(x) holds iff the letter at position x is a. The binary predicate S(x,y) denotes the successor relation on positions, and < is, as usual, its transitive closure. Example 1. Let A = {a,b} be the alphabet described by φA def = ∀x. Qa(x)∨Qb(x), which will be an additional conjunct below, not explicitly mentioned. – φ1 def = ∃x∃y. S(x,y)∧Qa(x)∧Qa(y) is a B(Σ1)[S] formula defining L1 = A∗aaA∗. – φ2 def = ∃x∃y. Qa(x)∧Qa(y)∧∀z. (x < z ⊃ y ≤ z) is a Σ2[<] formula defining L1. – Let φ3 def = (∀x. f irst(x) ⊃ Qa(x))∧ (∀x. last(x) ⊃ Qb(x))∧ (∀x,y. ((x < y)∧Qa(x)∧Qa(y) ⊃ ∃z. x < z∧ z < y∧Qb(z)))∧ (∀x,y. ((x < y)∧Qb(x)∧Qb(y) ⊃ ∃z. x < z∧ z < y∧Qa(z))) . Then, φ3 is a Π2[<] formula defining the language L2 = (ab)∗. ⊓⊔ More recently, Thérien and Wilke [TW98] showed that the 2-variable fragment FO2[<] [Mor75] (where only two variables occur, quantified any number of times), is expressively equivalent to the unambiguous languages and variety DA of Schützenberger [Sch76,TT02] and the subset ∆2[<] in the dot-depth hierarchy. Etessami, Vardi and Wilke [EVW02] identified the unary temporal logic LTL[F,P] and Schwentick, Thérien and Vollmer [STV02] identified partially-ordered 2-way deterministic finite automata (these are also called linear [LT00]) as equivalent formalisms. In [LPS08], we added to these correspondences a “deterministic” interval temporal logic called UITL. The papers [TW98,EVW02] also characterized FO2[<,S], which can define languages not definable in the logic FO2[<] such as those in Example 1. For a detailed study of these logics, see the recent papers of Weis and Immerman [WI07], and of Kufleitner and Weil [KW09]. PO2DFA ≡UL ≡ DA ≡ FO2[<] ≡ ∆2[<] ≡ LTL[F,P] ≡ UITL DA∗D ≡ FO2[<,S] ≡ ∆2[<,S] ≡ LTL[F,P,X,Y] It is clear that ∆2[<,S] ⊆ ∆3[<] since successor can be defined using < and one quantifier. In this paper we provide an automaton characterization and an interval logic characterization for this class of languages, and we separate it from B(Σ2)[<], the languages defined by the boolean combination of Σ2[<] formulae. This also shows that FO2[<,S] is a proper subset of ∆3[<], as diagrammatically depicted below.
منابع مشابه
Around Dot-Depth One
The dot-depth hierarchy is a classification of star-free languages. It is related to the quantifier alternation hierarchy of first-order logic over finite words. We consider fragments of languages with dot-depth 1/2 and dot-depth 1 obtained by prohibiting the specification of prefixes or suffixes. As it turns out, these language classes are in one-to-one correspondence with fragments of existen...
متن کاملSome results on the dot-depth hierarchy
In this paper we pursue the study of the decidability of the dot-depth hierarchy. We give an effective lower bound for the dot-depth of an aperiodic monoid. The main tool for this is the study of a certain operation on varieties of finite monoids in terms of Mal’cev product. We also prove the equality of two decidable varieties which were known to contain all dot-depth two monoids. Finally, we ...
متن کاملConnectionist Models Summer School
Binocular depth perception, or stereopsis, depends on matching corresponding points in two images taken from two vantage points. In random-dot stereograms the features to be matched are individual pixels. We have used the recurrent backpropagation learning algorithm of Pineda (1987) to construct network models with lateral and feedback connections that can solve the correspondence problem for r...
متن کاملLearning to see random-dot stereograms.
In the present study some specific properties of the learning effects reported for random-dot stereograms are examined. In experiment 1 the retinal position-specific learning effect was reproduced and in a follow-up experiment it was shown that the position specificity of learning can be accounted for by selective visual attention. In experiments 2 and 3 evidence was obtained that suggests that...
متن کاملThe efficiency of depth discrimination for non-transparent and transparent stereoscopic surfaces
The perception of transparency in binocular vision presents a challenge for any model of stereopsis. We investigate here how well human observers cope with stereo transparency by comparing their efficiency between transparent and opaque depth judgments. In two experiments, the efficiency measure was computed relative to an ideal observer to take into account the larger correspondence ambiguity ...
متن کاملTemporal properties of disparity processing revealed by dynamic random-dot stereograms.
In studies of the temporal flexibility of the stereoscopic system, it has been suggested that two different processes of binocular depth perception could be responsible for the flexibility: tolerance for interocular delays and temporal integration of correlation. None has investigated the relationship between tolerance for delays and temporal integration mechanisms and none has revealed which m...
متن کامل