According to Gross and Yellen (2004),[2] the algorithm can be traced back to Kleene (1956).[3] A presentation of the algorithm in the case of deterministic finite automata (DFAs) is given in Hopcroft and Ullman (1979).[4] The presentation of the algorithm for NFAs below follows Gross and Yellen (2004).[2]
the sets Rk ij of all strings that take M from state qi to qj without going through any state numbered higher than k.
Here, "going through a state" means entering and leaving it, so both i and j may be higher than k, but no intermediate state may.
Each set Rk ij is represented by a regular expression; the algorithm computes them step by step for k = -1, 0, ..., n. Since there is no state numbered higher than n, the regular expression Rn 0j represents the set of all strings that take M from its start stateq0 to qj. If F = { q1,...,qf } is the set of accept states, the regular expressionRn 01 | ... | Rn 0f represents the language accepted by M.
The initial regular expressions, for k = -1, are computed as follows for i≠j:
R−1 ij = a1 | ... | am where qj ∈ δ(qi,a1), ..., qj ∈ δ(qi,am)
and as follows for i=j:
R−1 ii = a1 | ... | am | ε where qi ∈ δ(qi,a1), ..., qi ∈ δ(qi,am)
In other words, R−1 ij mentions all letters that label a transition from i to j, and we also include ε in the case where i=j.
After that, in each step the expressions Rk ij are computed from the previous ones by
Rk ij = Rk-1 ik (Rk-1 kk)*Rk-1 kj | Rk-1 ij
Another way to understand the operation of the algorithm is as an "elimination method", where the states from 0 to n are successively removed: when state k is removed, the regular expression Rk-1 ij, which describes the words that label a path from state i>k to state j>k, is rewritten into Rk ij so as to take into account the possibility of going via the "eliminated" state k.
By induction on k, it can be shown that the length[5] of each expression Rk ij is at most 1/3(4k+1(6s+7) - 4) symbols, where s denotes the number of characters in Σ.
Therefore, the length of the regular expression representing the language accepted by M is at most 1/3(4n+1(6s+7)f - f - 3) symbols, where f denotes the number of final states.
This exponential blowup is inevitable, because there exist families of DFAs for which any equivalent regular expression must be of exponential size.[6]
In practice, the size of the regular expression obtained by running the algorithm can be very different depending on the order in which the states are considered by the procedure, i.e., the order in which they are numbered from 0 to n.
Example
Example DFA given to Kleene's algorithm
The automaton shown in the picture can be described as M = (Q, Σ, δ, q0, F) with
the set of states Q = { q0, q1, q2 },
the input alphabet Σ = { a, b },
the transition function δ with δ(q0,a)=q0, δ(q0,b)=q1, δ(q1,a)=q2, δ(q1,b)=q1, δ(q2,a)=q1, and δ(q2,b)=q1,
the start state q0, and
set of accept states F = { q1 }.
Kleene's algorithm computes the initial regular expressions as
R−1 00
= a | ε
R−1 01
= b
R−1 02
= ∅
R−1 10
= ∅
R−1 11
= b | ε
R−1 12
= a
R−1 20
= ∅
R−1 21
= a | b
R−1 22
= ε
After that, the Rk ij are computed from the Rk-1 ij step by step for k = 0, 1, 2.
Kleene algebra equalities are used to simplify the regular expressions as much as possible.
Step 0
R0 00
= R−1 00 (R−1 00)*R−1 00 | R−1 00
= (a | ε)
(a | ε)*
(a | ε)
| a | ε
= a*
R0 01
= R−1 00 (R−1 00)*R−1 01 | R−1 01
= (a | ε)
(a | ε)*
b
| b
= a*b
R0 02
= R−1 00 (R−1 00)*R−1 02 | R−1 02
= (a | ε)
(a | ε)*
∅
| ∅
= ∅
R0 10
= R−1 10 (R−1 00)*R−1 00 | R−1 10
= ∅
(a | ε)*
(a | ε)
| ∅
= ∅
R0 11
= R−1 10 (R−1 00)*R−1 01 | R−1 11
= ∅
(a | ε)*
b
| b | ε
= b | ε
R0 12
= R−1 10 (R−1 00)*R−1 02 | R−1 12
= ∅
(a | ε)*
∅
| a
= a
R0 20
= R−1 20 (R−1 00)*R−1 00 | R−1 20
= ∅
(a | ε)*
(a | ε)
| ∅
= ∅
R0 21
= R−1 20 (R−1 00)*R−1 01 | R−1 21
= ∅
(a | ε)*
b
| a | b
= a | b
R0 22
= R−1 20 (R−1 00)*R−1 02 | R−1 22
= ∅
(a | ε)*
∅
| ε
= ε
Step 1
R1 00
= R0 01 (R0 11)*R0 10 | R0 00
= a*b
(b | ε)*
∅
| a*
= a*
R1 01
= R0 01 (R0 11)*R0 11 | R0 01
= a*b
(b | ε)*
(b | ε)
| a*b
= a*b*b
R1 02
= R0 01 (R0 11)*R0 12 | R0 02
= a*b
(b | ε)*
a
| ∅
= a*b*ba
R1 10
= R0 11 (R0 11)*R0 10 | R0 10
= (b | ε)
(b | ε)*
∅
| ∅
= ∅
R1 11
= R0 11 (R0 11)*R0 11 | R0 11
= (b | ε)
(b | ε)*
(b | ε)
| b | ε
= b*
R1 12
= R0 11 (R0 11)*R0 12 | R0 12
= (b | ε)
(b | ε)*
a
| a
= b*a
R1 20
= R0 21 (R0 11)*R0 10 | R0 20
= (a | b)
(b | ε)*
∅
| ∅
= ∅
R1 21
= R0 21 (R0 11)*R0 11 | R0 21
= (a | b)
(b | ε)*
(b | ε)
| a | b
= (a | b) b*
R1 22
= R0 21 (R0 11)*R0 12 | R0 22
= (a | b)
(b | ε)*
a
| ε
= (a | b) b*a | ε
Step 2
R2 00
= R1 02 (R1 22)*R1 20 | R1 00
= a*b*ba
((a|b)b*a | ε)*
∅
| a*
= a*
R2 01
= R1 02 (R1 22)*R1 21 | R1 01
= a*b*ba
((a|b)b*a | ε)*
(a|b)b*
| a*b*b
= a*b (a (a | b) | b)*
R2 02
= R1 02 (R1 22)*R1 22 | R1 02
= a*b*ba
((a|b)b*a | ε)*
((a|b)b*a | ε)
| a*b*ba
= a*b*b (a (a | b) b*)*a
R2 10
= R1 12 (R1 22)*R1 20 | R1 10
= b*a
((a|b)b*a | ε)*
∅
| ∅
= ∅
R2 11
= R1 12 (R1 22)*R1 21 | R1 11
= b*a
((a|b)b*a | ε)*
(a|b)b*
| b*
= (a (a | b) | b)*
R2 12
= R1 12 (R1 22)*R1 22 | R1 12
= b*a
((a|b)b*a | ε)*
((a|b)b*a | ε)
| b*a
= (a (a | b) | b)*a
R2 20
= R1 22 (R1 22)*R1 20 | R1 20
= ((a|b)b*a | ε)
((a|b)b*a | ε)*
∅
| ∅
= ∅
R2 21
= R1 22 (R1 22)*R1 21 | R1 21
= ((a|b)b*a | ε)
((a|b)b*a | ε)*
(a|b)b*
| (a | b) b*
= (a | b) (a (a | b) | b)*
R2 22
= R1 22 (R1 22)*R1 22 | R1 22
= ((a|b)b*a | ε)
((a|b)b*a | ε)*
((a|b)b*a | ε)
| (a | b) b*a | ε
= ((a | b) b*a)*
Since q0 is the start state and q1 is the only accept state, the regular expression R2 01 denotes the set of all strings accepted by the automaton.
^McNaughton, R.; Yamada, H. (March 1960). "Regular Expressions and State Graphs for Automata". IRE Transactions on Electronic Computers. EC-9 (1): 39–47. doi:10.1109/TEC.1960.5221603. ISSN0367-9950.
^ abJonathan L. Gross and Jay Yellen, ed. (2004). Handbook of Graph Theory. Discrete Mathematics and it Applications. CRC Press. ISBN1-58488-090-2. Here: sect.2.1, remark R13 on p.65
^More precisely, the number of regular-expression symbols, "ai", "ε", "|", "*", "·"; not counting parentheses.
^Gruber, Hermann; Holzer, Markus (2008). "Finite Automata, Digraph Connectivity, and Regular Expression Size". In Aceto, Luca; Damgård, Ivan; Goldberg, Leslie Ann; Halldórsson, Magnús M.; Ingólfsdóttir, Anna; Walukiewicz, Igor (eds.). Automata, Languages and Programming. Lecture Notes in Computer Science. Vol. 5126. Springer Berlin Heidelberg. pp. 39–50. doi:10.1007/978-3-540-70583-3_4. ISBN9783540705833. S2CID10975422.. Theorem 16.