On the Algebraic Bases of Polyzetas

Basic Information

• Paper ID: 2510.13295
• Title: On the Algebraic Bases of Polyzetas
• Author: V. Hoang Ngoc Minh (University of Lille, France)
• Classification: math.CO (Combinatorics)
• Publication Date: October 15, 2025 (arXiv preprint)
• Paper Link: https://arxiv.org/abs/2510.13295

Abstract

This paper constructs two confluent rewriting systems on non-commutative polynomials, connecting two algebraic structures of polyzetas by identifying local coordinates (of the second kind) of the ζ-morphism graph as (shuffle or quasi-shuffle) characteristics. In each system, the left-hand side of rewriting rules corresponds to the leading monomial of the relevant homogeneous weight polynomial, while the right-hand side is canonically represented over the Q-algebra generated by irreducible terms, which encode the algebraic basis of the Q-algebra of polyzetas. These polynomials are completely ordered lexicographically and generate the kernel of the ζ-morphism, implying that the Q-free algebra of polyzetas is graded, irreducible polyzetas are transcendental, algebraically independent over Q, and consequently π² is Q-algebraically independent over odd zeta values (as is π itself).

Research Background and Motivation

Problem Background

1. Historical study of multiple zeta values: Multiple zeta values (MZV) $\zeta(s_1,...,s_r) = \sum_{n_1>...>n_r>0} \frac{1}{n_1^{s_1}...n_r^{s_r}}$, studied since Euler, are important objects in number theory
2. Algebraic structure problem: Understanding linear relationships and algebraic structures among these special values
3. Zagier's conjecture: Concerning the dimension $d_k$ of the Q-module generated by MZVs of weight k, satisfying the recurrence relation $d_k = d_{k-3} + d_{k-2}$ (k≥4)

Research Motivation

1. Need for symbolic methods: Compared to numerical methods such as the LLL algorithm, more rigorous algebraic methods are needed to determine linear relationships among MZVs
2. Transcendence problems: Proving that certain polyzetas are transcendental and algebraically independent
3. Unified framework: Establishing a bridge connecting shuffle and quasi-shuffle algebraic structures

Limitations of Existing Methods

1. Numerical methods: The LLL algorithm only provides "approximate certainty" results
2. Partial theories: Lack of a complete algebraic framework for handling all weight cases
3. Computational complexity: Calculations become difficult for high-weight cases

Core Contributions

1. Construction of two confluent rewriting systems: Establishing critical-pair-free rewriting systems $(Q1_{Y^*} \oplus (Y \setminus \{y_1\})Q\langle Y \rangle, R_Y^{irr})$ and $(Q1_{X^*} \oplus x_0Q\langle X \rangle x_1, R_X^{irr})$ on non-commutative polynomial rings
2. Establishment of Abel-type theorems: Proving limit relationships between non-commutative generating series of polylogarithms and harmonic sums
3. Determination of the kernel of ζ-morphism: Providing shuffle ideal $R_X$ and quasi-shuffle ideal $R_Y$ as the kernel of the ζ-morphism
4. Proof of gradedness of polyzeta algebra: Demonstrating that the Q-algebra $(Z, \times, 1)$ is freely graded, generated by irreducible polyzetas
5. Establishment of transcendence results: Proving that irreducible polyzetas are transcendental and algebraically independent, particularly that π is algebraically independent over odd zeta values

Detailed Methodology

Problem Definition

Studying algebraic relationships among multiple zeta values $\zeta(s_1,...,s_r)$, where $(s_1,...,s_r) \in H_r = \{(s_1,...,s_r) \in \mathbb{N}_{\geq 1}^r, s_1 > 1\}$.

Theoretical Framework

1. Functions on Free Monoids

Through one-to-one correspondence: $x_0^{s_1-1}x_1...x_0^{s_r-1}x_1 \in X^*x_1 \xleftrightarrow{\pi_Y}{\pi_X} y_{s_1}...y_{s_r} \in Y^* \leftrightarrow (s_1,...,s_r) \in \mathbb{N}_{\geq 1}^*$

Viewing polylogarithms, harmonic sums, and polyzetas as functions on free monoids.

2. Shuffle and Quasi-shuffle Algebras

• Shuffle product: $xu \shuffle yv = x(u \shuffle yv) + y(xu \shuffle v)$
• Quasi-shuffle product: $y_i u \star y_j v = y_i(u \star y_j v) + y_j(y_i u \star v) + y_{i+j}(u \star v)$

3. Group-like Series and MRS Form

Defining group-like series: $Z_{\shuffle} := \prod_{l \in Lyn_X \setminus X}^{\searrow} e^{\zeta(S_l)P_l}$$Z_{\star} := \prod_{l \in Lyn_Y \setminus \{y_1\}}^{\searrow} e^{\zeta(\Sigma_l)\Pi_l}$

Core Algorithm: LocalCoordinateIdentification

This algorithm discovers algebraic relationships among polyzetas by identifying local coordinates of group-like series:

1. Input: Weight upper bound p
2. Processing: For each $l \in Lyn_p^X$, identify coefficients in the equation $Z_\gamma = B(y_1)\pi_Y(Z_{\shuffle})$
3. Output: Rewriting rules $\Sigma_l \to \Upsilon_l$ and $S_l \to U_l$

Technical Innovations

1. Bridging equations: Using $Z_{\star} = B'(y_1)\pi_Y(Z_{\shuffle})$ to connect two algebraic structures
2. Lyndon word basis: Employing Lyndon words as purely transcendental bases to construct dual bases
3. Graded structure: Maintaining weight gradedness to ensure confluence of the rewriting system

Experimental Setup

Theoretical Verification

The paper is primarily theoretical research, with concrete calculations verified up to weight 12:

1. Weight range: Systematically analyzing all polyzetas of weights 2 through 12
2. Computational tools: Using symbolic computation to verify algebraic relationships
3. Comparative verification: Comparing with known results from Zagier's conjecture

Specific Example Analysis

The paper provides detailed computational tables demonstrating:

• Concrete forms of rewriting rules
• Identification of irreducible terms
• Verification of algebraic relationships

Experimental Results

Main Results

1. Successful Construction of Rewriting Systems

• Construction of two critical-pair-free confluent rewriting systems
• Determination of irreducible term sets $L_{X,\leq 12}^{irr}$ and $L_{Y,\leq 12}^{irr}$

2. Verification of Zagier's Conjecture

Up to weight 12, verification of: $Z_{X,\leq 12}^{irr} = \{\zeta(S_{x_0x_1}), \zeta(S_{x_0^2x_1}), \zeta(S_{x_0^4x_1}), \zeta(S_{x_0^6x_1}), \zeta(S_{x_0x_1^2x_0x_1^4}), \zeta(S_{x_0^8x_1}), ...\}$

The dimension sequence conforms to predictions from Zagier's conjecture.

3. Transcendence Results

Proving that:

• $\zeta(2q+1)/\pi^{2p} \notin \mathbb{Q}$ and $\zeta(2p+1)/\zeta(2q+1) \notin \mathbb{Q}$ (p≠q)
• π is algebraically independent over odd zeta values
• Irreducible polyzetas are transcendental

Concrete Numerical Examples

Rewriting rules for weight 4:

ζ(Σ_{y_4}) → (2/5)ζ(Σ_{y_2})²
ζ(Σ_{y_3y_1}) → (3/10)ζ(Σ_{y_2})²  
ζ(Σ_{y_2y_1²}) → (2/3)ζ(Σ_{y_2})²

Irreducible terms for weight 6:

ζ(Σ_{y_5}) is irreducible
ζ(Σ_{y_6}) → (8/35)ζ(Σ_{y_2})³

Related Work

Historical Development

1. Euler (1775): Earliest study of double zeta values, discovering fundamental identities
2. Nielsen (1906): Extending Euler's results, introducing more general identities
3. Zagier (1994): Proposing the conjecture on MZV dimensions
4. Modern developments: LLL algorithm, research on shuffle/stuffle relations

Advantages of This Work

1. Complete symbolization: Avoiding approximation inherent in numerical methods
2. Unified framework: Simultaneously handling shuffle and quasi-shuffle structures
3. Constructive proofs: Providing concrete algorithms and computational methods

Conclusions and Discussion

Main Conclusions

1. Algebraic structure: The Q-algebra of polyzetas is freely graded
2. Transcendence: Irreducible polyzetas are transcendental and algebraically independent
3. Computational methods: Providing systematic algorithms to determine algebraic relationships

Limitations

1. Computational complexity: Exponential growth of computation for high-weight cases
2. Implementation completeness: The LocalCoordinateIdentification algorithm is not fully implemented
3. Generality: Some results are only verified up to weight 12

Future Directions

1. Algorithm optimization: Improving computational efficiency for high-weight cases
2. Complete implementation: Fully implementing the algorithm
3. Generalization: Extending to more general L-function values

In-Depth Evaluation

Strengths

1. Theoretical depth: Providing a rigorous algebraic framework
2. Methodological innovation: Novel application of rewriting systems in this field
3. Important results: Addressing fundamental problems in polyzeta theory
4. Computational concreteness: Providing numerous concrete computational examples

Weaknesses

1. Complex exposition: Abundant symbols and definitions make reading difficult
2. Incomplete implementation: Core algorithm not fully implemented
3. Limited verification scope: Main results verified only for finite weights

Impact

1. Theoretical contribution: Providing new tools for MZV theory
2. Computational methods: Offering systematic methods for related computations
3. Transcendence theory: Important contributions to transcendental number theory

Applicable Scenarios

1. Number theory research: Particularly studies of special values
2. Algebraic geometry: Research involving periods
3. Symbolic computation: Implementation and application of related algorithms

References

The paper cites 33 important references, covering:

• Classical literature: Original works by Euler and Nielsen
• Modern theory: Important contributions by Zagier, Cartier, and others
• Computational methods: Numerical research by Borwein and colleagues
• Algebraic theory: Combinatorial algebra work by Reutenauer and others

Overall Assessment: This is an important paper in multiple zeta value theory, providing systematic methods for handling algebraic relationships of polyzetas. While technically complex, it possesses high theoretical value.