This paper introduces a new Composition-Diamond lemma for dialgebras using a more flexible ordering, establishing an equivalence between Gr"obner-Shirshov bases and linear bases, and providing practical normal form methods.
Contribution
It presents a more general and convenient Composition-Diamond lemma for dialgebras, ensuring the uniqueness of reduced Gr"obner-Shirshov bases and improving normal form computations.
Findings
01
New Composition-Diamond lemma with arbitrary monomial-center ordering
02
Equivalence of Gr"obner-Shirshov basis and linear basis conditions
03
Method for normal forms of elements in disemigroups
Abstract
Let Di⟨X⟩ be the free dialgebra over a field generated by a set X. Let S be a monic subset of Di⟨X⟩. A Composition-Diamond lemma for dialgebras is firstly established by Bokut, Chen and Liu in 2010 \cite{Di} which claims that if (i) S is a Gr\"{o}bner-Shirshov basis in Di⟨X⟩, then (ii) the set of S-irreducible words is a linear basis of the quotient dialgebra Di⟨X∣S⟩, but not conversely. Such a lemma based on a fixed ordering on normal diwords of Di⟨X⟩ and special definition of composition trivial modulo S. In this paper, by introducing an arbitrary monomial-center ordering and the usual definition of composition trivial modulo S, we give a new Composition-Diamond lemma for dialgebras which makes the conditions (i) and (ii) equivalent. We show that every ideal of Di⟨X⟩ has a…
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Full text
A new Composition-Diamond lemma for dialgebras111Supported by the NNSF of China (11571121) and the Science and Technology Program of Guangzhou (201605121833161).
Guangliang Zhang and
Yuqun
Chen222Corresponding author.
School of Mathematical Sciences, South China Normal
University
Abstract: Let Di⟨X⟩ be the free dialgebra over a field generated by a set X.
Let S be a monic subset of Di⟨X⟩. A Composition-Diamond lemma for dialgebras is
firstly established by Bokut, Chen and Liu in 2010 [6]
which claims that if (i) S is a Gröbner-Shirshov basis in Di⟨X⟩, then (ii) the
set of S-irreducible words is a linear basis of the quotient dialgebra Di⟨X∣S⟩,
but not conversely.
Such a lemma based on a fixed ordering on normal diwords of Di⟨X⟩ and special definition of composition trivial modulo S.
In this paper, by
introducing an arbitrary monomial-center ordering and the usual definition of composition trivial modulo S, we give a new Composition-Diamond lemma for dialgebras which makes the conditions (i) and (ii) equivalent. We show that every ideal of Di⟨X⟩ has a unique reduced Gröbner-Shirshov basis.
The new lemma is more useful and convenient than the one in [6].
As applications, we give a method to find normal forms of elements of an arbitrary disemigroup, in particular, A.V. Zhuchok’s (2010) and Y.V. Zhuchok’s (2015) normal forms of the free commutative disemigroups and the free abelian disemigroups, and normal forms of the free left (right) commutative disemigroups.
The notion of a dialgebra (disemigroup) was introduced by Loday [20] and
investigated in many papers (see, for example, [6, 17, 19, 20, 23, 26, 27, 28]).
Loday [20] constructed a free dialgebra and the universal enveloping dialgebra for a Leibniz algebra.
Bokut, Chen and Liu [6] established Gröbner-Shirshov bases theory for dialgebras.
Pozhidaev [23] studied the connection of Rota-Baxter algebras and dialgebras with associative bar-unity.
Kolesnikov [19] proved recently that each dialgebra may be obtained in turn from an associative conformal algebra.
Analogues of some notions of the functional analysis were defined on dialgebras in [17]. A.V. Zhuchok [26] and Y.V. Zhuchok [28] constructed the free commutative disemigroup and the free abelian disemigroup respectively. Various free disemigroups were introduced by A.V. Zhuchok in a survey paper [27].
Gröbner bases and Gröbner-Shirshov bases were invented independently by A.I. Shirshov
for ideals of free (commutative, anti-commutative) non-associative algebras [24, 25],
free Lie algebras [25] and implicitly free associative algebras [25] (see also [2, 3]), by
H. Hironaka [18] for ideals of the power series algebras (both formal and convergent),
and by B. Buchberger [12] for ideals of the polynomial algebras. Gröbner bases and
Gröbner-Shirshov bases theories have been proved to be very useful in different branches
of mathematics, including commutative algebra and combinatorial algebra. It is a powerful
tool to solve the following classical problems: normal form; word problem; conjugacy
problem; rewriting system; automaton; embedding theorem; PBW theorem; extension;
homology; growth function; Dehn function; complexity; etc. See, for example, the books
[1, 11, 13, 14, 15, 16] and the surveys [4, 5, 7, 8, 9, 10].
In Gröbner-Shirshov bases theory for a category of algebras, a key part is to establish
“Composition-Diamond lemma” for such algebras. The name “Composition-Diamond
lemma” combines the Neuman Diamond Lemma [22], the Shirshov Composition Lemma [24]
and the Bergman Diamond Lemma [2]. A Composition-Diamond lemma for
dialgebras was firstly given
by Bokut, Chen and Liu in 2010 [6].
Let Di⟨X⟩ be the free dialgebra over a field k generated by a well-ordered set X and X+ the free semigroup generated by X without the unit. With the notation as in [6], for any u=x1⋯xm⋯xn∈X+,
[TABLE]
is called a normal diword on X. The set [X+]ω of all normal diwords on X is a linear basis of Di⟨X⟩.
Let [X+]ω be a well-ordered set, S⊂Di⟨X⟩ a monic subset
of polynomials and Id(S) be the ideal of Di⟨X⟩ generated by S. A normal diword [u]n is
said to be S-irreducible if [u]n is not equal to the leading monomial of any normal S-diword. Let Irr(S) be
the set of all S-irreducible diwords. Consider the following statements:
(i) The set S is a Gröbner-Shirshov basis in Di⟨X⟩.
(ii) The set Irr(S) is a k-basis of the quotient dialgebra Di⟨X∣S⟩:=Di⟨X⟩/Id(S).
In [6], it is shown that (i)⇒(ii) but (ii)⇏(i).
Their proof of the above result based on a fixed ordering on [X+]ω and special definition of composition trivial modulo S.
In this paper, for an arbitrary monomial ordering on X+, we introduce a so-called monomial-center ordering on [X+]ω and give a new Composition-Diamond lemma for dialgebras
which makes the two conditions above equivalent, see Theorem 3.18. Comparing with the corresponding result in [6], the new lemma will be more useful and convenient when one calculates a Gröbner-Shirshov basis in Di⟨X⟩. We show that with a monomial-center ordering, every ideal of Di⟨X⟩ has a unique reduced Gröbner-Shirshov basis.
As applications, we give a method to find normal forms of elements of an arbitrary disemigroup.
In particular, we give short proofs of A.V. Zhuchok [26] and Y.V. Zhuchok’s [28] results on normal forms of elements of the free commutative disemigroup and the free abelian disemigroup generated by a set X, respectively.
Moreover, we give Gröbner-Shirshov bases for some dialgebras and disemigroups,
and obtain normal forms of elements of them.
The paper is organized as follows. In section 2, we review the free dialgebra Di⟨X⟩ over a field k generated by X.
In section 3, by introducing a monomial-center ordering on [X+]ω, normal S-diwords and compositions, we give a new Composition-Diamond lemma for dialgebras which makes the conditions (i) and (ii) mentioned before equivalent. In section 4, Gröbner-Shirshov bases theory for dirings is introduced, which may find an R-basis for some disemigroup-dirings over an associative ring R. In section 5, some applications are given.
2 Preliminaries
Throughout the paper, we fix a field k.
Z+ stands for the set of positive integers.
For a nonempty set X, we define the following notations:
X∗: the set of all associative words on X including the empty word, i.e. the free monoid generated by X.
X+: the set of all nonempty associative words on X, i.e. the free semigroup generated by X without the unit.
⌊X+⌋:={⌊xi1xi2⋯xin⌋∣i1,…,in∈I,i1≤i2≤⋯≤in,n∈Z+}, the set of all nonempty commutative associative words on X, where X={xi∣i∈I} is a total-ordered set.
[X+]ω:={[u]m∣u∈X+,m∈Z+,1≤m≤∣u∣}, the set of all associative normal diwords on X, following the notation in [6], where ∣u∣ is the number of letters in u (the length of u).
⌊X+⌋ω:={⌊u⌋m∣⌊u⌋∈⌊X+⌋,m∈Z+,1≤m≤∣u∣}, the set of all commutative normal diwords on X.
For u∈X+,[u]m is called an associative diword, while ⌊u⌋m is called a commutative diword. For example, if u=x2x1x2x1∈X+,x1<x2, then ⌊u⌋=⌊x1x1x2x2⌋,[u]3=x2x1x2˙x1,⌊u⌋3=⌊x1x1x2x2⌋3=x1x1x2˙x2.
⌊X+⌋1:={⌊u⌋1∣⌊u⌋∈⌊X+⌋}.
⌊X+⌋2−2:={⌊v⌋2∣⌊v⌋∈⌊X+⌋,∣v∣=2}.
Di⟨X⟩: the free dialgebra over a field k generated by X.
DiR⟨X⟩: the free diring over an associative ring R generated by X.
Disgp⟨X⟩=[X+]ω: the free disemigroup generated by X.
Di[X]: the free commutative dialgebra over a field k generated by X.
Disgp[X]=⌊X+⌋1∪⌊X+⌋2−2: the free commutative disemigroup generated by X.
Definition 2.1
*([21])
An associative dialgebra *(dialgebra for short) is ak-moduleDequipped with twok-linear maps
[TABLE]
where* ⊢and⊣are associative and satisfy the following identities:*
[TABLE]
for all* a,b,c∈D.*
A dialgebra* (D,⊢,⊣)is commutative if both * ⊢and⊣are commutative.
Write
[TABLE]
where ∣u∣ is the number of letters in u. For any h=[u]m∈[X+]ω,
we call u the associative word of h, and
m, denoted by p(h), the position of center of h.
For example, if u=x1x2⋯xn∈X+, xi∈X,h=[u]m,1≤m≤n, then p(h)=m and with the notation as in [6],
[TABLE]
A word [u]m∈[X+]ω is called a normal diword.
Let Di⟨X⟩ be the free k-module with a k-basis [X+]ω.
For any [u]m,[v]n∈[X+]ω, define
[TABLE]
and extend them linearly to Di⟨X⟩.
It is well known from [21] that Di⟨X⟩ is the free dialgebra generated by X.
Let X be a well-ordered set. We define the deg-lex ordering on X+ by the following:
for u=xi1xi2⋯xin,v=xj1xj2⋯xjm∈X+, where each xil,xjt∈X,
[TABLE]
An ordering > on X+ is said to be monomial if > is a well ordering and for any u,v,w∈X+,
[TABLE]
Clearly, the deg-lex ordering is monomial.
3 A new Composition-Diamond lemma
Let > be a monomial ordering on X+.
We define the monomial-center ordering>d on [X+]ω as follows. For any [u]m,[v]n∈[X+]ω,
[TABLE]
In particular, if > is the deg-lex ordering on X+, we call the ordering defined by (\refequ0) the deg-lex-center ordering on [X+]ω.
For simplicity of notation, we write > instead of >d when no confusion can arise.
It is clear that a monomial-center ordering is a well ordering on [X+]ω. Such an ordering plays an important role in this paper. Here and subsequently, the monomial-center ordering on [X+]ω will be used, unless otherwise stated.
For convenience we assume that [u]m>0 for any [u]m∈[X+]ω.
For any nonzero polynomial f∈Di⟨X⟩, let us denote f be the leading monomial of f with respect to the ordering >,
lt(f) the leading term of f, lc(f) the coefficient of f and f the associative word of f.
f is called monic if lc(f)=1. For any nonempty subset S of Di⟨X⟩,
S is* monic* if s is monic for all s∈S.
Definition 3.1
A nonzero polynomial* f∈Di⟨X⟩is strong iff>rf, whererf:=f−lt(f).*
It is easy to check that
> on [X+]ω is monomial in the following sense:
[TABLE]
where [u]m,[v]n,[w]l∈[X+]ω.
From this it follows that
Lemma 3.2
Let 0=f∈Di⟨X⟩ and [u]m∈[X+]ω. Then
[TABLE]
In particular, if f is strong, then ([u]m⊣f)=[u]m⊣f and (f⊢[u]m)=f⊢[u]m.
Example 3.3
Let X={x1,x2,x3}, x1>x2>x3, Chark=2,3 and > be the deg-lex-center ordering on [X+]ω.
Let f=2[x1x2x3]3−2[x1x2x3]2+3[x1x3]2. Then
[TABLE]
The polynomial f is not strong since f=x1x2x3=rf. Of course, rf is strong. It is easy to check that
[TABLE]
Here and subsequently, S denotes a monic subset of Di⟨X⟩ unless otherwise stated.
By an S-diwordg we mean
a normal diword on X∪S with only one occurrence of s∈S. If this is the case and
[TABLE]
where 1≤k≤n,xil∈X,1≤l≤n, then we also call g an s-diword. For simplicity, we denote the s-diword of the form (3) by (asb), where a,b∈X∗, s∈S.
Definition 3.4
An* S-diword(3)is called a normal S-diword if eitherk=morsis strong.*
Note that if (asb) is a normal S-diword, then (asb)=[asb]l for some l∈P([asb]), where
[TABLE]
If this is so, we denote the normal S-diword (asb) by [asb]l and also call [asb]l a normal s-diword.
In what follows, to simplify notation, we let
[TABLE]
where [u]m,[v]n∈[X+]ω, f∈Di⟨X⟩.
The lemma below follows from Definition 3.4.
Lemma 3.5
Let (asb) be an s-diword and [u]m,[v]n∈[X+]ω.
Then (asb)=[asb]l
if and only if [u]m⊢(asb)⊣[v]n=[uasbv]∣u∣+l, where u,v may be empty.
Definition 3.6
Let* f,gbe monic polynomials inDi⟨X⟩.*
If* fis not strong, then we callx⊣fthe composition of left multiplication offfor allx∈Xandf⊢[u]∣u∣the composition of right multiplication offfor allu∈X+.*
2. 2)
Suppose that* w=f=agbfor somea,b∈X∗and(agb)is a normalg-diword.*
2.1 Ifp(f)∈P([agb]),
then we call
[TABLE]
the* composition of inclusion offandg.*
2.2 Ifp(f)∈/P([agb])and bothfandgare strong,
then for anyx∈Xwe call
[TABLE]
the* composition of left multiplicative inclusion offandg, and*
[TABLE]
the* composition of right multiplicative inclusion offandg.*
3. 3)
Suppose that there exists a* w=fb=agfor somea,b∈X∗such that∣f∣+∣g∣>∣w∣, (fb)is a normal f-diword
and(ag)is a normalg-diword.*
3.1 IfP([fb])∩P([ag])=∅,
then for anym∈P([fb])∩P([ag])we call
[TABLE]
the* composition of intersection offandg.*
3.2 IfP([fb])∩P([ag])=∅and bothfandgare strong,
then for anyx∈Xwe call
[TABLE]
the* composition of left multiplicative intersection offandg, and*
[TABLE]
the* composition of right multiplicative intersection offandg.*
For any composition (f,g)[u]n mentioned above, we call [u]n the ambiguity of f and g.
Definition 3.7
Let* Sbe a monic subset ofDi⟨X⟩and[w]m∈[X+]ω.
A polynomialh∈Di⟨X⟩is trivial modulo S((S,[w]m), resp.), denoted by*
[TABLE]
if* h=∑αi[aisibi]mi, where eachαi∈k,ai,bi∈X∗,si∈Sand[aisibi]mi≤h([aisibi]mi<[w]m, resp.).*
A monic set* Sis called a Gröbner-Shirshov basis inDi⟨X⟩if any
composition of polynomials inSis trivial moduloS.*
A monic set S is said to be closed under the composition of left (right, resp.) multiplication
if all left (right, resp.) multiplication compositions of elements of S are trivial modulo S.
We set
Irr(S):={[u]n∈[X+]ω∣[u]n=[asb]m for any normal S-diword [asb]m }.
Remark 3.8
The definition of a Gröbner-Shirshov basis in [6]
is different from the Definition 3.7.* In [6], the definition of a Gröbner-Shirshov basis is based on a fixed ordering on [X+]ω.Comparing with [6], we have different definitions of the following: ordering of normal diwords; normal S-diword;
compositions of left and right multiplication, multiplicative inclusion and multiplicative intersection; and composition to be trivial.*
In calculating a Gröbner-Shirshov basis in Di⟨X⟩, the following example shows that our method is more convenient than the one of [6]. In all examples of this section, we let > be the deg-lex-center ordering on [X+]ω, where X is a well-ordered set.
Example 3.9
Let D=Di⟨X∣S⟩.
If S⊆[X+]ω, then it is easy to check that S is a Gröbner-Shirshov basis.
But the result is not true in the sense of [6].
For example, let X={x1,x2,x3}, x1>x2>x3, and D=Di⟨X∣[x1x2]2⟩.
Then S={[x1x2]2} is a Gröbner-Shirshov basis.
Applying Theorem 3.18 we conclude that
[TABLE]
is a linear basis of D. Let S1={[x1x2]2,[xx1x2]1∣x∈X}.
In the sense of [6], S1 is a Gröbner-Shirshov basis, but S is not.
However, Irr(S1) in the sense of [6] is the same as the set Irr(S).
Lemma 3.10
Let S be closed under the composition of left multiplication and f∈S.
If f is not strong, then for any [u]1∈[X+]ω,
[u]1⊣f≡0mod(S).
Proof. The proof follows by induction on (uf,∣u∣).
If ∣u∣=1, then the result holds.
Assume that ∣u∣≥2 and [u]1=[vx]1, v∈X+, x∈X. Then
[u]1⊣f=[v]1⊣(x⊣f) is a linear combination of S-diwords of the form
[v]1⊣[asb]m,
where s∈S and [asb]m≤(x⊣f).
It follows that ([v]1⊣[asb]m)≤[v]1⊣[asb]m≤[v]1⊣(x⊣f)=([u]1⊣f) and asb≤xf.
If s is strong, then [v]1⊣[asb]m is already a normal S-diword, and we have done.
Suppose that s is not strong. If a is empty, then [v]1⊣[asb]m=([v]1⊣s)⊣[b]1 and (vs,∣v∣)<(uf,∣u∣).
If a is not empty, then [v]1⊣[asb]m=([va]1⊣s)⊣[b]1 and m=∣a∣+p(s)>1.
Since [asb]m≤[xf]1, we have asb<xf and (vas,∣va∣)<(uf,∣u∣).
By induction, [v]1⊣[asb]m is a linear combination of S-diwords of the form
[cs′d]n⊣[b]1, where s′∈S and [cs′d]n≤([va]1⊣s).
By Lemma 3.5, [cs′d]n⊣[b]1 is a normal S-diword, and
[cs′d]n⊣[b]1≤([va]1⊣s)⊣[b]1=([v]1⊣[asb]m)≤([u]1⊣f).
□
Remark 3.11
The following example shows that
Lemma 3.10 is not true if we replace* “x⊣f”, “[u]1⊣f” by “f⊢x”, “f⊢[u]∣u∣” respectively.*
Example 3.12
Let X={x1,x2},x1>x2, Chark=2 and S={f,g,h}, where f=[x1x2]2+[x1x2]1,
g=[x1x2x1]3−21[x1x2x1]2−21[x1x2x1]1,
h=[x1x2x2]3−21[x1x2x2]2−21[x1x2x2]1.
Clearly, f,g and h are not strong. We check at once that g⊢xi=0, h⊢xi=0, i=1,2, and
[TABLE]
However, f⊢[x1x1]2 is not trivial modulo S.
Lemma 3.13
Let S be closed under the compositions of left and right multiplication.
Then for any normal S-diword [asb]m and [u]n∈[X+]ω,
[TABLE]
Moreover, if asb<w, w∈X+,
then
[TABLE]
Proof.
We prove only the results for the case [u]n⊣[asb]m. The proof of the another case is similar.
If s is strong, then [u]n⊣[asb]m is a normal S-diword.
Assume that s is not strong. Note that [u]n=[u1]∣u1∣⊢[u2]1, u1,u2∈X∗,∣u1∣=n−1.
Then
[TABLE]
By Lemma 3.10,
[u]n⊣[asb]m is a linear combination of S-diwords of the form
[u1]∣u1∣⊢[cs′d]l⊣[b]1,
where s′∈S and [cs′d]l≤([u2a]1⊣s).
By Lemma 3.5, [u1]∣u1∣⊢[cs′d]l⊣[b]1 is a normal S-diword,
and
[TABLE]
If asb<w, w∈X+, then uasb<uw and ([u]n⊣[asb]m)≤[uasb]n<[uw]n.
□
The lemma below follows from Lemma 3.13, Definitions 3.6 and 3.7 immediately.
Lemma 3.14
Let S be a Gröbner-Shirshov basis in Di⟨X⟩, f,g strong polynomials in S, x∈X, [u]n∈[X+]ω. Then the following statements hold.
(i)
If w=fb=ag for some a,b∈X∗ such that ∣f∣+∣g∣>∣w∣,
then
[TABLE]
Moreover,
[TABLE]
2. (ii)
If w=f=agb for some a,b∈X∗,
then
[TABLE]
Moreover,
[TABLE]
Lemma 3.15
Let S be closed under the compositions of left and right multiplication.
Then for any S-diword (asb), (asb) has an expression:
[TABLE]
where each αi∈k,si∈S,ai,bi∈X∗, and [aisibi]mi≤(asb).
Proof. We may assume that
[TABLE]
If k=m or s is strong, then (asb) is a normal S-diword and the result holds.
Suppose that k=m and s is not strong. Then
[TABLE]
Clearly, [a]∣a∣⊢s and s⊣[b]1 are normal S-diwords.
In both cases, the result follows from Lemma 3.13. □
Lemma 3.16
Let S be a monic subset of Di⟨X⟩.
Then for any nonzero polynomial f∈Di⟨X⟩,
[TABLE]
where each [ui]ni∈Irr(S),αi,βj∈k,aj,bj∈X∗,sj∈S, [ui]ni≤f and
[ajsjbj]mj≤f.
Proof.
Let f=lc(f)f+rf. If f∈Irr(S),
then take [u]n=f and f1=f−lc(f)[u]n. If f∈/Irr(S),
then f=[asb]m for some normal S-diword [asb]m and take f1=f−lc(f)[asb]m.
In both cases, we have f1<f and the result
follows from induction on f. □
Lemma 3.17
Let S be a Gröbner-Shirshov basis in Di⟨X⟩ and [a1s1b1]m1,[a2s2b2]m2 normal S-diwords.
If [w]m=[a1s1b1]m1=[a2s2b2]m2, then
[TABLE]
Proof. Since [w]m=[a1s1b1]m1=[a2s2b2]m2,
it follows that w=a1s1b1=a2s2b2 and m=m1=m2.
Here we need consider three cases:
[TABLE]
For Case 1, we may assume that s1 is at the left of s2, i.e.
b1=as2b2 and a2=a1s1a, here a may be empty.
Then
[TABLE]
Let s1=s1+∑βi[ui]ni, s2=s2+∑βj′[vj]lj.
Here we have to discuss five cases:
[TABLE]
We first give the proof for Case 1.1. The same proof remains valid for Cases 1.3 and 1.5.
Since 1≤m1≤∣a1∣, we have
[TABLE]
As s1,s2 are strong we have
[TABLE]
[TABLE]
and
ui<s1,vj<s2. It follows that
[TABLE]
We proceed to show Case 1.2. Similar proof applies to Case 1.4.
Since m1=∣a1∣+p(s1), we have
[TABLE]
It is clear that
[TABLE]
As s1 is strong we also have
[TABLE]
It follows that
[TABLE]
We now turn to Case 2, and may assume that s2 is a subword of s1, say,
w′=s1=as2b. Then a2=a1a, b2=bb1 and
[TABLE]
Here we also should discuss five cases:
[TABLE]
In Cases 2.1 and 2.2, let a1=a1′x and in Cases 2.4 and 2.5, let b1=yb1′, where a1′,b1′∈X∗, x,y∈X.
Then
[TABLE]
By Lemmas 3.5 and 3.14, (4) and (8) are linear combinations of normal S-diwords
with leading monomials less than [a1′xw′b1]m1=[a1w′yb1′]m1=[w]m1.
Applying Lemmas 3.5, 3.14 and using the fact that S is a Gröbner-Shirshov basis,
the same conclusion can be drawn for (5), (6) and (7).
For Case 3, we may assume that s1 is at the left of s2, i.e.
a2=a1a, b1=bb2, and w′=s1b=as2.
Then
[TABLE]
Here we continue to discuss five cases:
[TABLE]
In Cases 3.1 and 3.2, let a1=a1′x and in Cases 3.4 and 3.5, let b1=yb1′, where a1′,b1′∈X∗, x,y∈X.
Then
[TABLE]
By Lemmas 3.5 and 3.14, (9) and (13) are linear combinations of normal S-diwords
with leading monomials less than [a1′xw′b2]m1=[a1w′yb2′]m1=[w]m1.
Applying Lemmas 3.5, 3.14 and using the fact that S is a Gröbner-Shirshov basis,
the same conclusion can be drawn for (10), (11) and (12).
□
Theorem 3.18
(Composition-Diamond lemma for dialgebras) Let S be a monic subset of Di⟨X⟩ ,
> a monomial-center ordering on [X+]ω and
Id(S) the ideal of Di⟨X⟩ generated by S. Then the following statements are equivalent.
(i)
S* is a Gröbner-Shirshov basis in Di⟨X⟩.*
2. (ii)
f∈Id(S)⇒f=[asb]m* for some normal S-diword [asb]m.*
3. (iii)
Irr(S)={[u]n∈[X+]ω∣[u]n=[asb]m\mboxforanynormalS\mbox−diword[asb]m}*
is a k-basis of the quotient dialgebra Di⟨X∣S⟩:=Di⟨X⟩/Id(S).*
Proof.(i)⇒(ii). Let 0=f∈Id(S). Then by
Lemma 3.15f has an expression
[TABLE]
where each αi∈k,ai,bi∈X∗,si∈S. Write
[wi]mi=[aisibi]mi=[aisibi]mi,i=1,2,⋯.
We may assume without loss of generality that
[TABLE]
The proof follows by induction on ([w1]m1,l). If l=1, then
f=[a1s1b1]m1=[a1s1b1]m1 and
the result holds. Suppose that l≥2.
Then
[TABLE]
By Lemma 3.17, we can rewrite the first two summands of (14) in the form
[TABLE]
where each [cjsj′dj]nj is a normal S-diword and [cjsj′dj]nj<[w1]m1.
Thus the result follows from induction on ([w1]m1,l).
(ii)⇒(iii). By Lemma 3.16, the set Irr(S) generates
Di⟨X∣S⟩ as a linear space. On the other hand, suppose that
h=∑αi[ui]li=0 in Di⟨X∣S⟩, where
each αi∈k, [ui]li∈Irr(S). This means that
h∈Id(S).
Then all αi must be equal to zero. Otherwise, h=[uj]lj for some j which contradicts (ii).
(iii)⇒(i). Suppose that h is a composition
of elements of S. Clearly, h∈Id(S). By Lemma 3.16,
[TABLE]
where each [ui]ni∈Irr(S),αi,βj∈k,aj,bj∈X∗,sj∈S, and [ui]ni≤h,[ajsjbj]mj≤h.
Then ∑iαi[ui]ni∈Id(S).
By (iii), we have αi=0 and
h≡0mod(S).
□
Remark 3.19
In [6], a Composition-Diamond lemma for dialgebras is established and claims that*
(i)⇒(iii), but not conversely. The reason is that the definitions of
a Gröbner-Shirshov basis inDi⟨X⟩are different, see Remark 3.8.*
Shirshov algorithm If a monic subset S⊂Di⟨X⟩ is not a Gröbner-Shirshov basis then one can add to S all
nontrivial compositions. Continuing this process
repeatedly, we finally obtain a Gröbner-Shirshov basis Scomp that contains S
and generates the same ideal, Id(Scomp)=Id(S).
Definition 3.20
A Gröbner-Shirshov basis* SinDi⟨X⟩is minimal if for anys∈S, s∈Irr(S\{s}).
A Gröbner-Shirshov basisSinDi⟨X⟩is reduced if for anys∈S, supp(s)⊆Irr(S\{s}), where*
Suppose* Iis an ideal ofDi⟨X⟩andI=Id(S). IfSis a (reduced) Gröbner-Shirshov basis inDi⟨X⟩, then we also callSis a (reduced) Gröbner-Shirshov basis for the ideal Ior for the quotient dialgebra Di⟨X⟩/I.*
For associative algebras and polynomial algebras, it is known that every ideal has a unique reduced Gröbner-Shirshov basis.
This result is still true for dialgebras.
Lemma 3.21
Let I be an ideal of Di⟨X⟩ and S a Gröbner-Shirshov basis for I. For any T⊆S,
if Irr(T)=Irr(S) then T is also a Gröbner-Shirshov basis for I.
Proof. For any f∈I, since Irr(T)=Irr(S) and S a Gröbner-Shirshov basis for I=Id(S), we have, by Theorem 3.18, f=[asb]m=[cgd]m for some s∈S,g∈T,a,b,c,d∈X∗. Thus,
f1=f−lc(f)[cgd]m∈I and f1<f. By induction on f, f is a linear combination of normal T-diwords, i.e. f∈Id(T). This shows that I=Id(T). Now the result follows from Theorem 3.18. □
Let S be a subset of Di⟨X⟩ and [w]m∈[X+]ω. We set
[TABLE]
Theorem 3.22
Let I be an ideal of Di⟨X⟩ and > a monomial-center ordering on [X+]ω.
Then there is a unique reduced Gröbner-Shirshov basis for I.
Proof.
It is clear that there is a Gröbner-Shirshov basis S for I, for example, we may take S={lc(f)−1f∣0=f∈I}.
For each [w]m∈S, we choose a polynomial f[w]m in S such that f[w]m=[w]m. Write
[TABLE]
Noting that I⊇S⊇S0 and I=S=S0, we have Irr(S0)=Irr(S)=[X+]ω\S.
By Lemma 3.21, S0 is a Gröbner-Shirshov basis for I.
Moreover, we may assume that for any s∈S0,
[TABLE]
i.e.
supp(s−s)⊆[X+]ω\S0.
Indeed, if supp(s−s)∩S0=∅ for some s∈S0, then set
[u]n=max{supp(s−s)∩S0} and there is an f∈S0 such that f=[u]n.
Note that s>[u]n=f and s−f=s. Replace s by s−f in S0.
Then supp(s−f−s−f)∩S0=∅ or max{supp(s−f−s−f)∩S0}<[u]n.
Since > is a well ordering on [X+]ω, this process will terminate.
Note that for any [w]m∈S0, there exists a unique f∈S0 such that [w]m=f.
Set min{S0}=s0 with s0∈S0. Define
Ss0:={s0}.
Suppose that f∈S0,s0<f and Sg has been defined for any g∈S0 with g<f. Define
[TABLE]
Let
[TABLE]
Then for any f∈S0,f∈S1⇔f∈Irr(S<fˉ)⇔f∈Sf.
We first claim that Irr(S1)=Irr(S0). Noting that S1⊆S0, it suffices to show that Irr(S1)⊆Irr(S0). Assume that there is [w]m∈[X+]ω such that [w]m∈Irr(S1) and [w]m∈/Irr(S0).
Since S0=I, it follows that [w]m=f for some f∈S0\S1. If f∈Irr(S<f) then f∈Sf⊆S1, a contradiction. If f∈Irr(S<f) then
f=[asb]m for some s∈S<f⊆S1,a,b∈X∗.
This implies that f∈Irr(S1), a contradiction. Therefore, Irr(S1)=Irr(S0). Now by Lemma 3.21, S1 is a Gröbner-Shirshov basis for I.
If f,g∈S1,f=g,f=[agb]m, then g<f,g∈Sg⊆S<f which implies f∈Irr(S<f) and f∈S1, a contradiction. This shows that S1 is a minimal Gröbner-Shirshov basis for I.
By (15), for any s∈S1, supp(s)⊆Irr(S1\{s}) which means S1 is a reduced Gröbner-Shirshov basis for I.
This shows that I has a reduced Gröbner-Shirshov basis S1.
Suppose that T is an arbitrary reduced Gröbner-Shirshov basis for I. Let s0=minS1 and
r0=minT, where s0∈S1,r0∈T. By Theorem 3.18,
s0=[a′r′b′]p≥r′≥r0 for some r′∈T,a′,b′∈X∗.
Similarly, r0≥s0. Then r0=s0. We say that r0=s0.
Otherwise, 0=r0−s0∈I. We apply the above argument again, with replace s0 by r0−s0,
to obtain that r0>r0−s0≥r′′≥r0 for some r′′∈T, a contradiction.
As both T and S1 are reduced Gröbner-Shirshov bases, we have S1s0={s0}={r0}=Tr0.
Given any [w]m∈S1∪T with [w]m>r0.
Assume that S1<[w]m=T<[w]m.
To prove T=S1, it is sufficient to show that S1[w]m⊆T[w]m.
For any s∈S1[w]m, we can see that
s=[c′rd′]q≥r for some r∈T,c′,d′∈X∗.
Now, we claim that [w]m=s=r. Otherwise, [w]m=s>r.
Then r∈T<[w]m=S1<[w]m and r∈S1\{s}.
But s=[c′rd′]q, which contradicts the fact that S1 is a reduced Gröbner-Shirshov basis.
We next claim that s=r∈T[w]m. If s=r, then 0=s−r∈I.
By Theorem 3.18, s−r=[ar1b]n=[cs1d]n for some
r1∈T,s1∈S1,a,b,c,d∈X∗ with
r1,s1≤s−r<s=r.
This means that s1∈S1\{s} and r1∈T\{r}.
Noting that s−r∈supp(s)∪supp(r), we may assume that s−r∈supp(s).
As S1 is a reduced Gröbner-Shirshov basis, we have s−r∈Irr(S1\{s}),
which contradicts the fact that s−r=[cs1d]n, where s1∈S1\{s}.
Thus s=r. This shows that S1[w]m⊆T[w]m.
□
Remark 3.23
For associative algebras and polynomial algebras, it is known that every Gröbner-Shirshov basis for an ideal can be reduced to a
reduced Gröbner-Shirshov basis for the ideal.
Unfortunately, for dialgebras, this is not the case.**
The following example shows that generally, a Gröbner-Shirshov basis S in Di⟨X⟩ may not be reduced to
a minimal Gröbner-Shirshov basis for I=Id(S).
Example 3.24
Let X={x}, Chark=2,3 and S={f,g,h,p},
where
[TABLE]
Then S
is a Gröbner-Shirshov basis in Di⟨X⟩
and S can not be reduced to
a minimal Gröbner-Shirshov basis for I=Id(S).
Proof. We first show that all compositions in S are trivial.
Compositions of left (right) multiplication.
All possible compositions of left (right) multiplication are ones related to g,h,p.
By noting that for any xn∈X+, we have
[TABLE]
[TABLE]
Compositions of inclusion and left (right) multiplicative inclusion.
We denote by, for example, “f∧g,[w]m” the composition of the polynomials of f and g with ambiguity [w]m.
By noting that in S,
[TABLE]
all possible of compositions of inclusion in S are:
[TABLE]
As g,h,p are not strong, there is no composition of left (right) multiplicative inclusion.
For f∧g, [w]m=[x4]4, we have
[TABLE]
For h∧g, [w]m=[x4]3, we have
[TABLE]
Compositions of intersection and left (right) multiplicative intersection.
By noting that in S,
[TABLE]
all possible ambiguities [w]m of compositions of intersection are:
[TABLE]
As g,h,p are not strong, there is no composition of left (right) multiplicative intersection.
It is easy to see that all the compositions of intersection are trivial modulo S.
This shows that S is a Gröbner-Shirshov basis.
Note that in S, f=[xg]4, [xg]4 is a normal g-diword, and f−[xg]4=21h. Since 2[x5]1=x⊣h is nontrivial modulo {g,h,p}, {g,h,p} is not a Gröbner-Shirshov basis. This implies that we can not drop f from S, i.e. S can not be reduced to
a minimal Gröbner-Shirshov basis.
□
4 Gröbner-Shirshov bases for dirings
In this section, by similar proofs of the above section, we introduce Gröbner-Shirshov bases for dirings, which may find an R-basis for some disemigroup-dirings over an associative ring R.
Definition 4.1
([21])
A disemigroup is a setDequipped with two maps
[TABLE]
where* ⊢and⊣are associative and satisfy the identities (1).*
Note that in [21, 26, 27, 28], such a disemigroup in the above definition is called a dimonoid.
It is well known from [21] that ([X+]ω,⊢,⊣) is the free disemigroup generated by X, where
for any [u]m,[v]n∈[X+]ω,
[TABLE]
Let us denote
[TABLE]
the free disemigroup generated by X.
Throughout this section, R is an associative ring with unit.
Definition 4.2
A* diring is a quaternary(T,+,⊢,⊣)such that both(T,+,⊢)and(T,+,⊣)are associative rings with the identities (1)
inT.*
Definition 4.3
Let* (D,⊢,⊣)be a disemigroup, Ran associative ring with unit andTthe free leftR-module withR-basisD. Then(T,+,⊢,⊣)is a diring with a natural way: for anyf=∑iriui,g=∑jrj′vj∈T,ri,rj′∈R,ui,vj∈D,*
[TABLE]
Such a diring, denoted by DiR(D), is called a* disemigroup-diring ofDoverR.*
We denote by* DiR⟨X⟩the disemigroup-diring ofDisgp⟨X⟩overRwhich is also called the free diring overRgenerated byX.
In particular, Dik⟨X⟩=Di⟨X⟩is the free dialgebragenerated byXwhenkis a field.*
As same as the proof of Theorem 3.18, we have the following Composition-Diamond lemma for dirings.
Theorem 4.4
(Composition-Diamond lemma for dirings) Let S be a monic subset of DiR⟨X⟩,
> a monomial-center ordering on [X+]ω and
Id(S) the ideal of DiR⟨X⟩ generated by S. Then the following statements are equivalent.
(i)
S* is a Gröbner-Shirshov basis in DiR⟨X⟩.*
2. (ii)
f∈Id(S)⇒f=[asb]m* for some normal S-diword [asb]m.*
3. (iii)
Irr(S)={[u]n∈[X+]ω∣[u]n=[asb]m* for any normal S-diword [asb]m}
is an R-basis of the quotient diring DiR⟨X∣S⟩:=DiR⟨X⟩/Id(S), i.e. DiR⟨X∣S⟩ is a free R-module with R-basis Irr(S).*
Remark 4.5
Shirshov algorithm does not work generally in* DiR⟨X⟩.*
5 Applications
In this section, by using our Theorem 3.18, we give a method to find normal forms of elements of an arbitrary disemigroup, in particular, we give normal forms of elements of free commutative disemigroups, free abelian disemigroups and free left (right) commutative disemigroups.
5.1 Normal forms of disemigroups
For an arbitrary disemigroup D, D has an expression
[TABLE]
for some set X and S⊆[X+]ω×[X+]ω, where ρ(S) is the congruence on ([X+]ω,⊢,⊣) generated by S.
It is natural to ask how to find normal forms of elements of disemigroup Disgp⟨X∣S⟩?
Let > be a monomial-center ordering on [X+]ω and S={([ui]mi,[vi]ni)∣[ui]mi>[vi]ni,i∈I}. Consider the dialgebra Di⟨X∣S⟩, where S={[ui]mi−[vi]ni∣i∈I}.
By Shirshov algorithm, we have a Gröbner-Shirshov basis Scomp in Di⟨X⟩ and Id(Scomp)=Id(S). It is clear that each element in Scomp is of the form [u]m−[v]n, where [u]m>[v]n,[u]m,[v]n∈[X+]ω. Let
[TABLE]
Then σ is obviously a dialgebra isomorphism. Since Irr(Scomp) is a linear basis of Di⟨X∣S⟩, we have σ(Irr(Scomp)) is a linear basis of
Dik([X+]ω/ρ(S)) which shows that Irr(Scomp) is exactly normal
forms of elements of the disemigroup Disgp⟨X∣S⟩.
Therefore, we have the following theorem.
Theorem 5.1
Let > be a monomial-center ordering on [X+]ω and D=Disgp⟨X∣S⟩, where S={([ui]mi,[vi]ni)∣[ui]mi>[vi]ni,i∈I} is a subset of [X+]ω×[X+]ω. Then Irr(Scomp) is a set of normal forms of elements of the disemigroup
Disgp⟨X∣S⟩.
From now on, let > be the deg-lex-center ordering on [X+]ω, where X is a well-ordered set.
5.2 Normal forms of free commutative disemigroups
The commutative disemigroups are introduced and the free commutative disemigroup generated by a set is constructed by [26]. In this subsection, we give another approach to normal forms of elements of a free commutative disemigroup.
Definition 5.2
([26])
A disemigroup(D,⊢,⊣)is commutative if both⊢and⊣are commutative.
Let Di[X] be the free commutative dialgebra generated by a set X and T be the subset of Di⟨X⟩
consisting of the following polynomials:
[TABLE]
where [u]m,[v]n∈[X+]ω.
Then Di[X]=Di⟨X∣T⟩ and Disgp[X]=Disgp⟨X∣T⟩ is the free commutative disemigroup generated by X.
Let X={xi∣i∈I} be a total-ordered set,
[TABLE]
the set of all nonempty commutative associative words on X and
[TABLE]
the set of all commutative normal diwords on X.
For u∈X+,[u]m is called an associative diword, while ⌊u⌋m is called a commutative diword. For example, if u=x2x1x2x1∈X+,x1<x2, then ⌊u⌋=⌊x1x1x2x2⌋,[u]3=x2x1x2˙x1,⌊u⌋3=⌊x1x1x2x2⌋3=x1x1x2˙x2.
Proposition 5.3
Let X={xi∣i∈I} be a well-ordered set. Then
(i)
Di[X]=Di⟨X∣S⟩, where S consists of the following
polynomials:
[TABLE]
2. (ii)
S* is a Gröbner-Shirshov basis in Di⟨X⟩.*
3. (iii)
The set
⌊X+⌋1∪⌊X+⌋2−2
is a k-basis of the free commutative dialgebra Di[X],
where
[TABLE]
Proof.(i) We only need to prove that the polynomials in S are trivial modulo T and the polynomials in T are trivial modulo S.
It is clear that
[TABLE]
where xi,xj∈X,v∈X+,∣v∣≥2. Suppose that v=xj1⋯xjn⋯xjl∈X+, n≥2,l>2.
If n<l, then v=v1xjnv2 for some v1,v2∈X+ and
[TABLE]
If n=l, then v=xj1v′xjn for some v′∈X+ and
[TABLE]
It is easily seen that
[TABLE]
where x,y∈X.
Suppose that [u]m,[v]n∈[X+]ω with ∣uv∣>2.
[TABLE]
(ii) It is easy to check that all possible compositions of left (right) multiplication in S are equal to zero.
For any composition of (f,g)[w]m in S, note that −rf,−rg∈[X+]ω, ∣w∣≥3,
[w]m=[afb]m=[cgd]m and
⌊w⌋1=⌊arfb⌋1=⌊crgd⌋1,
where f=f+rf,g=g+rg, a,b,c,d∈X∗. It follows that
[TABLE]
Then all the compositions in S are trivial. We have proved (ii).
From Theorem 3.18, Lemma 3.21 and Proposition 5.3, it follows that
Corollary 5.4
Let W be a set consisting of the following
polynomials:
[TABLE]
Then
W is the reduced Gröbner-Shirshov basis for the free commutative dialgebra Di[X].
From Theorem 5.1 and Proposition 5.3, it follows that
Corollary 5.5
([26, Theorem 3])
Disgp[X]=(⌊X+⌋1∪⌊X+⌋2−2,⊢,⊣) is the free commutative disemigroup generated by X,
where the operations ⊢ and ⊣ are as follows: for any
x,x′∈X,⌊u⌋p1,⌊v⌋p2∈⌊X+⌋1∪⌊X+⌋2−2 with ∣u∣∣v∣>1,
[TABLE]
5.3 Normal forms of free abelian disemigroups
The concept of abelian disemigroups is introduced and the free abelian disemigroup generated by a set is constructed by [28]. In this subsection, we give another approach to normal forms of elements of a free abelian disemigroup.
Definition 5.6
([28])
A disemigroup(D,⊢,⊣)is abelian ifa⊢b=b⊣afor alla,b∈D.
Let X be an arbitrary set and T the subset of [X+]ω×[X+]ω
consisting of the following:
[TABLE]
where [u]m,[v]n∈[X+]ω. Then Disgp⟨X∣T⟩ is the free abelian disemigroup generated by X.
Let X={xi∣i∈I} be a total-ordered set.
Suppose that u=xj1xj2⋯xjn∈X+ and ⌊u⌋=⌊xi1xi2⋯xin⌋,
where xi1,xi2,⋯,xin is the reordering of
xj1,xj2,⋯,xjn such that xi1≤xi2≤⋯≤xin.
Define
[TABLE]
For example, if u=x1x2x1x2 with x1<x2, then ⌊u⌋=⌊x1x1x2x2⌋, cont(u)={x1,x2}, L(u)={1,2,3,4},
ρu(2)=x2, λ⌊u⌋(x2)=3, τu(2)=3.
For any u,v∈X+, it is easy to check that τ⌊u⌋(m)≤m for all m∈L(u)
and τuv(∣u∣+n)=τvu(n) for all n∈L(v).
Proposition 5.7
Let X={xi∣i∈I} be a well-ordered set,
T the subset of Di⟨X⟩
consisting of the following polynomials:
[u]m⊢[v]n−[v]n⊣[u]m, where [u]m,[v]n∈[X+]ω. Then
(i)
Di⟨X∣T⟩=Di⟨X∣S⟩, where S consists of the following
polynomials:
[TABLE]
2. (ii)
S* is a Gröbner-Shirshov basis in Di⟨X⟩.*
3. (iii)
The set
[TABLE]
is a k-basis of the free abelian dialgebra Di⟨X∣T⟩.
Proof.(i)
We only need to prove that the polynomials in S are trivial modulo T and the polynomials in T are trivial modulo S.
[TABLE]
On the other hand, it is easy to see that [u]m−⌊u⌋τu(m)≡0mod(T).
(ii) It is easy to check that all possible compositions of left (right) multiplication in S are equal to zero.
For any composition of (f,g)[w]m in S, note that −rf,−rg∈[X+]ω, ∣w∣≥3,
[w]m=[afb]m=[cgd]m and
⌊w⌋=⌊arfb⌋=⌊crgd⌋,
where f=f+rf,g=g+rg, a,b,c,d∈X∗. It follows that
[TABLE]
From the definition of composition in S we conclude that
ρw(m)=ρarfb(m1)=ρcrgd(m2).
Thus τarfb(m1)=λ⌊arfb⌋ρarfb(m1)=λ⌊crgd⌋ρcrgd(m2)=τcrgd(m2)
and
[TABLE]
Then all the compositions in S are trivial. We have proved (ii).
Then
(FAd(X),⊢,⊣) is the free abelian disemigroup generated by X,
where the operations ⊢ and ⊣ are as follows: for any ⌊u⌋t,⌊v⌋p∈FAd(X),
[TABLE]
5.4 Normal forms of free left (right) commutative disemigroups
Definition 5.10
A disemigroup* (D,⊢,⊣)is left (right) commutative
ifa⊣b⊣c=b⊣a⊣c,a⊢b⊢c=b⊢a⊢c(a⊢b⊢c=a⊢c⊢b,a⊣b⊣c=a⊣c⊣b)for alla,b,c∈D.*
Let X be an arbitrary set and T the subset of [X+]ω×[X+]ω,
where T consists of the following:
[TABLE]
where [u]m,[v]n,[w]l∈[X+]ω. Then Disgp⟨X∣T⟩ is the free left commutative disemigroup generated by X.
Proposition 5.11
Let X={xi∣i∈I} be a well-ordered set and T the subset of Di⟨X⟩
consisting of the following polynomials:
[TABLE]
where [u]m,[v]n,[w]l∈[X+]ω. Then
(i)
Di⟨X∣T⟩=Di⟨X∣S⟩, where S consists of the following
polynomials:
[TABLE]
2. (ii)
S* is a Gröbner-Shirshov basis in Di⟨X⟩.*
3. (iii)
The set
[TABLE]
is a k-basis of the dialgebra Di⟨X∣T⟩ and
normal forms of elements of the free left commutative disemigroup Disgp⟨X∣T⟩.
Proof.(i) We only need to prove that the polynomials in S are trivial modulo T and the polynomials in T are trivial modulo S.
[TABLE]
Suppose that [w]l=[w1]∣w1∣⊢[xw2]1=[w′y]l
where w1,w2,w′∈X∗,x,y∈X.
[TABLE]
[TABLE]
[TABLE]
(ii) It is easy to check that all possible compositions of left (right) multiplication in S are equal to zero.
For any composition of (f,g)[w]m in S, note that −rf,−rg∈[X+]ω, ∣w∣≥4,
[w]m=[afb]m=[cgd]m and
⌊w⌋=⌊arfb⌋=⌊crgd⌋,
where a,b,c,d∈X∗. It follows that
[TABLE]
If ∣w∣−m≤1, then m1=m2=m and ∣arfb∣−m1≤1, ∣crgd∣−m2≤1.
[TABLE]
where u,v∈X∗, x∈X, ∣u∣≥2,∣v∣≤1.
If ∣w∣−m>1, then m1≤m,m2≤m and ∣arfb∣−m1>1, ∣crgd∣−m2>1.
[TABLE]
where u,v∈X∗, x,y∈X, ∣v∣≥1.
Then all the compositions in S are trivial. We have proved (ii).
(iii) This part follows from Theorems 3.18 and 5.1.
□
From Theorem 3.18, Lemma 3.21 and Proposition 5.11, it follows that
Corollary 5.12
Let W be a set consisting of the following
polynomials:
[TABLE]
Then
W is the reduced Gröbner-Shirshov basis for the free left commutative dialgebra Di⟨X∣T⟩.
Analysis similar to that in the proof of Proposition 5.11 shows the following proposition.
Proposition 5.13
Let X={xi∣i∈I} be a well-ordered set and T′ the subset of Di⟨X⟩
consisting of the following polynomials:
[TABLE]
where [u]m,[v]n,[w]l∈[X+]ω. Then
(i)
Di⟨X∣T′⟩=Di⟨X∣S′⟩, where S′ consists of the following
polynomials:
[TABLE]
2. (ii)
S* is a Gröbner-Shirshov basis in Di⟨X⟩.*
3. (iii)
The set
[TABLE]
is a k-basis of the dialgebra Di⟨X∣T′⟩ and
normal forms of elements of the free right commutative disemigroup Disgp⟨X∣T′⟩.
Corollary 5.14
Let W′ be a set consisting of the following
polynomials:
[TABLE]
Then
W′ is the reduced Gröbner-Shirshov basis for the free right commutative dialgebra Di⟨X∣T′⟩.
Bibliography28
The reference list from the paper itself. Each links out to its DOI / PubMed record.
1[1] W. W. Adams, P. Loustaunau, An introduction to Gröbner bases, Graduate Studies in Mathematics, vol. 3, American Mathematical Society, 1994.
2[2] G. M. Bergman, The diamond lemma for ring theory, Adv. in Math. 29(2) (1978) 178–218.
3[3] L. A. Bokut, Imbeddings into simple associative algebras, Algebra i Logika 15(2) (1976) 117–142, 245.
4[4] L. A. Bokut, Y. Chen, Gröbner-Shirshov bases: Some new results, in: Proceedings of the Second International Congress in Algebra and Combinatorics, World Scientific, 2008, pp. 35–56.
5[5] L. A. Bokut, Y. Chen, Gröbner-Shirshov Bases and Shirshov Algorithm, Educational tutorial lecture notes, Novosibirsk State University, 2014.
6[6] L. A. Bokut, Y. Chen, C. Liu, Gröbner-Shirshov bases for dialgebras, Internat. J. Algebra Comput. 20(3) (2010) 391–415.
7[7] L. A. Bokut, Y. Chen, Gröbner-Shirshov bases and their calculation, Bull. Math. Sci. 4(3) (2014) 325–395.
8[8] L. A. Bokut, Y. Fong, V.-F. Ke, P. S. Kolesnikov, Gröbner and Gröbner-Shirshov bases in algebra, and conformal algebras, Fundam. Prikl. Mat. 6(3) (2000) 669–706.