local recompression word equations and beyond
play

Local recompression Word Equations and Beyond Artur Je Max Planck - PowerPoint PPT Presentation

Local recompression Word Equations and Beyond Artur Je Max Planck Institute for Informatics 21 June 2013 21 June 2013 1/32 Word Equations Definition Given equation U = V , where U , V ( X ) . Is there an assignment S : X


  1. Dealing with crossing pairs ab is a crossing pair There is X such that S ( X ) = bw and aX appears in U = V (or symmetric). replace X with bX (implicitly change solution S ( X ) = bw to S ( X ) = w ) If S ( X ) = ǫ then remove X . Lemma After performing this for all variables, ab is no longer crossing. 21 June 2013 12/32

  2. Dealing with crossing pairs ab is a crossing pair There is X such that S ( X ) = bw and aX appears in U = V (or symmetric). replace X with bX (implicitly change solution S ( X ) = bw to S ( X ) = w ) If S ( X ) = ǫ then remove X . Lemma After performing this for all variables, ab is no longer crossing. Compress the pair! 21 June 2013 12/32

  3. Example XbaYb = baaababbab for S ( X ) = baaa S ( Y ) = bba ab is a crossing pair 21 June 2013 13/32

  4. Example XbaYb = baaababbab for S ( X ) = baaa S ( Y ) = bba ab is a crossing pair replace X with Xa , Y with bYa (new solution: S ( X ) = baa , S ( Y ) = b ) XababYab = baaababbab for S ( X ) = baa S ( Y ) = b 21 June 2013 13/32

  5. Example XbaYb = baaababbab for S ( X ) = baaa S ( Y ) = bba ab is a crossing pair replace X with Xa , Y with bYa (new solution: S ( X ) = baa , S ( Y ) = b ) XababYab = baaababbab for S ( X ) = baa S ( Y ) = b ab is not longer crossing, we replace it by c XccY c = baaccbc for S ( X ) = baa S ( Y ) = b 21 June 2013 13/32

  6. Maximal blocks Definition ( maximal block of a ) When a ℓ appears in S ( U ) = S ( V ) and cannot be extended. Block appearance can be explicit, implicit or crossing. Letter a has crossing block if there is a crossing ℓ -block of a . 21 June 2013 14/32

  7. Maximal blocks Definition ( maximal block of a ) When a ℓ appears in S ( U ) = S ( V ) and cannot be extended. Block appearance can be explicit, implicit or crossing. Letter a has crossing block if there is a crossing ℓ -block of a . Equivalents of pairs. Compress them similarly. Pop whole prefixes/suffixes, not single letters 21 June 2013 14/32

  8. Maximal blocks Definition ( maximal block of a ) When a ℓ appears in S ( U ) = S ( V ) and cannot be extended. Block appearance can be explicit, implicit or crossing. Letter a has crossing block if there is a crossing ℓ -block of a . Equivalents of pairs. Compress them similarly. Pop whole prefixes/suffixes, not single letters Lemma (Length-minimal solutions) For maximal a ℓ block: ℓ ≤ 2 cn . 21 June 2013 14/32

  9. Blocks compression Definition (Crossing block) maximal block is crossing iff it is contained in S ( U ) ( S ( V ) ) but not in explicit words nor in any S ( X ) . 21 June 2013 15/32

  10. Blocks compression Definition (Crossing block) maximal block is crossing iff it is contained in S ( U ) ( S ( V ) ) but not in explicit words nor in any S ( X ) . When a has no crossing block 1: for all maximal blocks a ℓ of a do let a ℓ ∈ Σ be a unused letter 2: replace each explicit maximal a ℓ in U = V by a ℓ 3: 21 June 2013 15/32

  11. What about crossing blocks? Idea change the equation X defines a ℓ X wa r X : change it to w replace X in equation by a ℓ X Xa r X 21 June 2013 16/32

  12. What about crossing blocks? Idea change the equation X defines a ℓ X wa r X : change it to w replace X in equation by a ℓ X Xa r X CutPrefSuff ( a ) 1: for X ∈ X do guess and remove a -prefix a ℓ i and a -suffix a r X of S ( X ) 2: replace each X in rules bodies by a ℓ X Xa r X 3: 21 June 2013 16/32

  13. What about crossing blocks? Idea change the equation X defines a ℓ X wa r X : change it to w replace X in equation by a ℓ X Xa r X CutPrefSuff ( a ) 1: for X ∈ X do guess and remove a -prefix a ℓ i and a -suffix a r X of S ( X ) 2: replace each X in rules bodies by a ℓ X Xa r X 3: Lemma After CutPrefSuff ( a ) letter a has no crossing block. 21 June 2013 16/32

  14. What about crossing blocks? Idea change the equation X defines a ℓ X wa r X : change it to w replace X in equation by a ℓ X Xa r X CutPrefSuff ( a ) 1: for X ∈ X do guess and remove a -prefix a ℓ i and a -suffix a r X of S ( X ) 2: replace each X in rules bodies by a ℓ X Xa r X 3: Lemma After CutPrefSuff ( a ) letter a has no crossing block. So a ’s blocks can be easily compressed. 21 June 2013 16/32

  15. What about crossing blocks? Idea change the equation X defines a ℓ X wb r X : change it to w replace X in equation by a ℓ X Xb r X CutPrefSuff 1: for X ∈ X do let X begin with a and end with b 2: calculate and remove a -prefix a ℓ X and b -suffix b r X of X 3: replace each X in rules bodies by a ℓ X Xb r X 4: Lemma After CutPrefSuff no letter has a crossing block. So all blocks can be easily compressed. 21 June 2013 16/32

  16. Algorithm while U / ∈ Σ and V / ∈ Σ do L ← letters from U = V uncross the blocks for a ∈ L do compress a blocks 21 June 2013 17/32

  17. Algorithm while U / ∈ Σ and V / ∈ Σ do L ← letters from U = V uncross the blocks for a ∈ L do compress a blocks P ← noncrossing pairs of letters from U = V ⊲ Guess P ′ ← crossing pairs of letters from U = V ⊲ Guess, only O ( n ) for ab ∈ P do compress pair ab for ab ∈ P ′ do uncross and compress pair ab 21 June 2013 17/32

  18. Crucial property Theorem (Main property: shortens the solution) Let ab be a string in U = V or in S ( X ) (for a length-minimal S). At least one of a, b is compressed in one phase. 21 June 2013 18/32

  19. Crucial property Theorem (Main property: shortens the solution) Let ab be a string in U = V or in S ( X ) (for a length-minimal S). At least one of a, b is compressed in one phase. Proof. a = b By block compression. a � = b Pair compression tries to compress ab . Fails, when one was compressed already. 21 June 2013 18/32

  20. Crucial property Theorem (Main property: shortens the solution) Let ab be a string in U = V or in S ( X ) (for a length-minimal S). At least one of a, b is compressed in one phase. Proof. a = b By block compression. a � = b Pair compression tries to compress ab . Fails, when one was compressed already. Corollary (Running time) The algorithm has O ( log N ) phases. 21 June 2013 18/32

  21. Space consumption Corollary (Space consumption) The equation has length O ( n 2 ) . 21 June 2013 19/32

  22. Space consumption Corollary (Space consumption) The equation has length O ( n 2 ) . Proof. we introduce O ( n ) letters per uncrossing O ( n ) uncrossings in one phase: O ( n 2 ) new letters and we shorten it by a constant factor in each phase. | U ′ | + | V ′ | ≤ 2 3 ( | U | + | V | ) + cn 2 Gives quadratic upper bound on the whole equation. 21 June 2013 19/32

  23. Solution upper bound Idea Running time is at most ( cn 2 ) cn 2 . there are O ( log N ) phases So log N ∼ ( cn 2 ) cn 2 . 21 June 2013 20/32

  24. Solution upper bound Idea Running time is at most ( cn 2 ) cn 2 . there are O ( log N ) phases So log N ∼ ( cn 2 ) cn 2 . Lemma There are Ω( log N ) / poly ( n ) phases 21 June 2013 20/32

  25. Solution upper bound Idea Running time is at most ( cn 2 ) cn 2 . there are O ( log N ) phases So log N ∼ ( cn 2 ) cn 2 . Lemma There are Ω( log N ) / poly ( n ) phases Proof. We do not shorten too much (at most 2 cn letters into one). 21 June 2013 20/32

  26. Solution upper bound Idea Running time is at most ( cn 2 ) cn 2 . there are O ( log N ) phases So log N ∼ ( cn 2 ) cn 2 . Lemma There are Ω( log N ) / poly ( n ) phases Proof. We do not shorten too much (at most 2 cn letters into one). log N / poly ( n ) ≤ ( cn 2 ) cn 2 21 June 2013 20/32

  27. Linear space consumption Aim at O ( n ) space consumption O ( 1 ) pair-uncrossing per variable smarter block compression 21 June 2013 21/32

  28. Improving pair compression Partition of pairs Σ ℓ and Σ r are disjoint: we can compress pairs from Σ ℓ Σ r in parallel 21 June 2013 22/32

  29. Improving pair compression Partition of pairs Σ ℓ and Σ r are disjoint: we can compress pairs from Σ ℓ Σ r in parallel choose partition that covers many appearances: think of random partition, it covers half of pairs in equation only O ( 1 ) uncrossing per pair 21 June 2013 22/32

  30. Improving pair compression Partition of pairs Σ ℓ and Σ r are disjoint: we can compress pairs from Σ ℓ Σ r in parallel choose partition that covers many appearances: think of random partition, it covers half of pairs in equation only O ( 1 ) uncrossing per pair This given O ( n ) long equation. 21 June 2013 22/32

  31. Block compression Idea when we replace a blocks, only equality matters, not length pop a ℓ X and b r X from X but treat them as parameters 21 June 2013 23/32

  32. Block compression Idea when we replace a blocks, only equality matters, not length pop a ℓ X and b r X from X but treat them as parameters guess the equal blocks check if they can be equal replace them 21 June 2013 23/32

  33. Block compression Idea when we replace a blocks, only equality matters, not length pop a ℓ X and b r X from X but treat them as parameters guess the equal blocks check if they can be equal replace them Length of a block b a a a a b c c d a a a a a a a a a a b c c d a a a X Y X 21 June 2013 23/32

  34. Block compression Idea when we replace a blocks, only equality matters, not length pop a ℓ X and b r X from X but treat them as parameters guess the equal blocks check if they can be equal replace them Length of a block b a a a a b c c d a a a a a a a a a a b c c d a a a X Y X Linear combination of { ℓ X , r X } X ∈X and constants. 21 June 2013 23/32

  35. Verification Guessed equalities ⇐ ⇒ system of linear Diophantine equations in { ℓ X , r X } X ∈X 21 June 2013 24/32

  36. Verification Guessed equalities ⇐ ⇒ system of linear Diophantine equations in { ℓ X , r X } X ∈X has size proportional to equation – encode variables as in the equation – encode constants in unary 21 June 2013 24/32

  37. Verification Guessed equalities ⇐ ⇒ system of linear Diophantine equations in { ℓ X , r X } X ∈X has size proportional to equation – encode variables as in the equation – encode constants in unary can be verified in linear space (nondeterministically) – iteratively guess parity 21 June 2013 24/32

  38. Verification Guessed equalities ⇐ ⇒ system of linear Diophantine equations in { ℓ X , r X } X ∈X has size proportional to equation – encode variables as in the equation – encode constants in unary can be verified in linear space (nondeterministically) – iteratively guess parity Linear space. 21 June 2013 24/32

  39. Linear Space Equation is of length O ( n ) . each letter may be different, so O ( n log n ) bits want true linear space 21 June 2013 25/32

  40. Linear Space Equation is of length O ( n ) . each letter may be different, so O ( n log n ) bits want true linear space Tools special encoding of letters: represented by fragments of the original equation – letters representing only original letters: appropriate tree – those representing other letters: depend only on XwY , encode them like that 21 June 2013 25/32

  41. Linear Space Equation is of length O ( n ) . each letter may be different, so O ( n log n ) bits want true linear space Tools special encoding of letters: represented by fragments of the original equation – letters representing only original letters: appropriate tree – those representing other letters: depend only on XwY , encode them like that improve the pair compression (special pairing by Sakamoto) quite technical 21 June 2013 25/32

  42. Exponent of periodicity Definition ⇒ u k is a substring of w but u ′ k + 1 is not. per ( w ) = k ⇐ ⇒ a k is a substring of w but b k + 1 is not. per Σ ( w ) = k ⇐ 21 June 2013 26/32

  43. Exponent of periodicity Definition ⇒ u k is a substring of w but u ′ k + 1 is not. per ( w ) = k ⇐ ⇒ a k is a substring of w but b k + 1 is not. per Σ ( w ) = k ⇐ We do not fully use per ( S ( U )) , only per Σ ( S ( U )) . per Σ ( S ( U )) is the length of maximal block. 21 June 2013 26/32

  44. Exponent of periodicity Definition ⇒ u k is a substring of w but u ′ k + 1 is not. per ( w ) = k ⇐ ⇒ a k is a substring of w but b k + 1 is not. per Σ ( w ) = k ⇐ We do not fully use per ( S ( U )) , only per Σ ( S ( U )) . per Σ ( S ( U )) is the length of maximal block. Those are (components of) solution of a Diophantine system in { ℓ X , r X } X ∈X They are at most exponential (standard algebra and analysis). So per Σ ( S ( U )) is at most exponential. 21 June 2013 26/32

  45. Reducing per to per Σ Lemma Consider S ( U ) and S ( U ′ ) obtained by compression of pairs (or blocks). Let u k be a substring of S ( U ) . Then either 21 June 2013 27/32

  46. Reducing per to per Σ Lemma Consider S ( U ) and S ( U ′ ) obtained by compression of pairs (or blocks). Let u k be a substring of S ( U ) . Then either u ∈ a ∗ or u ′ k − 1 is a substring of S ′ ( U ′ ) (for some u ′ ) 21 June 2013 27/32

  47. Reducing per to per Σ Lemma Consider S ( U ) and S ( U ′ ) obtained by compression of pairs (or blocks). Let u k be a substring of S ( U ) . Then either u ∈ a ∗ or u ′ k − 1 is a substring of S ′ ( U ′ ) (for some u ′ ) Proof. If u k is not a block then compression does not affect u k too much, u ′ k − 1 can be chosen. 21 June 2013 27/32

  48. Reducing per to per Σ Lemma Consider S ( U ) and S ( U ′ ) obtained by compression of pairs (or blocks). Let u k be a substring of S ( U ) . Then either u ∈ a ∗ or u ′ k − 1 is a substring of S ′ ( U ′ ) (for some u ′ ) Proof. If u k is not a block then compression does not affect u k too much, u ′ k − 1 can be chosen. There are O (( cn ) cn ) compression steps. In each of them per ( U = V ) = per Σ ( U = V ) (exponential) or it drops by a constant. So per ( U = V ) is at most exponential. 21 June 2013 27/32

  49. Univariate equations Form of the equation A = B A 0 XA 1 . . . A k − 1 XA k = XB 1 . . . B k − 1 XB k , where A i , B i ∈ Σ ∗ , A 0 � = ǫ . 21 June 2013 28/32

  50. Univariate equations Form of the equation A = B A 0 XA 1 . . . A k − 1 XA k = XB 1 . . . B k − 1 XB k , where A i , B i ∈ Σ ∗ , A 0 � = ǫ . Nondeterminism dissappears only S ( X ) � = ǫ first (last) letter of S ( X ) is known S ( X ) ∈ a ∗ are easy to check; otherwise a -prefix of S ( X ) and A 0 have the same length 21 June 2013 28/32

  51. Univariate equations Form of the equation A = B A 0 XA 1 . . . A k − 1 XA k = XB 1 . . . B k − 1 XB k , where A i , B i ∈ Σ ∗ , A 0 � = ǫ . Nondeterminism dissappears only S ( X ) � = ǫ first (last) letter of S ( X ) is known S ( X ) ∈ a ∗ are easy to check; otherwise a -prefix of S ( X ) and A 0 have the same length Whenever we pop, we test some solution. 21 June 2013 28/32

  52. Generating all solutions We want a finite (graph-like) representation of all solutions. Not all solutions are length minimal. 21 June 2013 29/32

  53. Generating all solutions We want a finite (graph-like) representation of all solutions. Not all solutions are length minimal. Lemma (Plandowski) It is enough to consider minimal solutions. Minimal under homomorphism. 21 June 2013 29/32

  54. Generating all solutions We want a finite (graph-like) representation of all solutions. Not all solutions are length minimal. Lemma (Plandowski) It is enough to consider minimal solutions. Minimal under homomorphism. Lemma (Minimal solutions) If ab has an implicit appearance in S ( U ) for a minimal S then ab has crossing or explicit one. 21 June 2013 29/32

  55. Transforming solutions Definition (Transforming solutions) An operation changing U = V to U ′ = V ′ transforms solutions if we can associate an operator H with it such that when S ′ is a solution of U ′ = V ′ then S = H [ S ′ ] each S is of this form 21 June 2013 30/32

  56. Transforming solutions Definition (Transforming solutions) A nondeterministic operation changing U = V to U ′ = V ′ transforms solutions if we can associate a family of operators H depending on choices with it such that when S ′ is a solution of U ′ = V ′ then S = H [ S ′ ] for each H ∈ H each S is of this form for some nondeterministic choices and some H ∈ H 21 June 2013 30/32

  57. Transforming solutions Definition (Transforming solutions) An operation changing U = V to U ′ = V ′ transforms solutions if we can associate an operator H (depending on choices) with it such that when S ′ is a solution of U ′ = V ′ then S = H [ S ′ ] each S is of this form 21 June 2013 30/32

  58. Transforming solutions Definition (Transforming solutions) An operation changing U = V to U ′ = V ′ transforms solutions if we can associate an operator H (depending on choices) with it such that when S ′ is a solution of U ′ = V ′ then S = H [ S ′ ] each S is of this form Lemma All our operations transform solutions. Operators are easy to define (morphisms). 21 June 2013 30/32

  59. Representation of solutions Definition ( G ) nodes: equations, edges: U = V is transformed to U ′ = V ′ label: operator H transforming the solution 21 June 2013 31/32

  60. Representation of solutions Definition ( G ) nodes: equations, edges: U = V is transformed to U ′ = V ′ label: operator H transforming the solution trivial equations have simple solutions each solution corresponds to a path in G and vice-versa 21 June 2013 31/32

  61. Representation of solutions Definition ( G ) nodes: equations, edges: U = V is transformed to U ′ = V ′ label: operator H transforming the solution trivial equations have simple solutions each solution corresponds to a path in G and vice-versa Construction Verify the nodes’ existance. Verify edges’ existance. Labels are natural to deduce from the algorithm. PSPACE [Matching Plandowski’s construction] 21 June 2013 31/32

  62. Open questions, related research, etc. Also used for fully compressed membership problem for NFAs [in NP] fully compressed pattern matching [quadratic algorithm] approximation of the smallest grammar [simpler algorithm] . . . ? 21 June 2013 32/32

Recommend


More recommend