Multi-Dimensional Customization Modelling Based On Metagraph For Saas Multi-Tenant Applications

Software as a Service (SaaS) is a new software delivery model in which pre-built applications are delivered to customers as a service. SaaS providers aim to attract a large number of tenants (users) with minimal system modifications to meet economics of scale. To achieve this aim, SaaS applications have to be customizable to meet requirements of each tenant. However, due to the rapid growing of the SaaS, SaaS applications could have thousands of tenants with a huge number of ways to customize applications. Modularizing such customizations still is a highly complex task. Additionally, due to the big variation of requirements for tenants, no single customization model is appropriate for all tenants. In this paper, we propose a multi-dimensional customization model based on metagraph. The proposed mode addresses the modelling variability among tenants, describes customizations and their relationships, and guarantees the correctness of SaaS customizations made by tenants.

💡 Research Summary

The paper addresses a fundamental challenge in modern Software‑as‑a‑Service (SaaS) platforms: how to support thousands of tenants, each with highly individualized customization requirements, without incurring prohibitive development and maintenance costs. Traditional customization approaches treat the problem as a single‑dimensional feature selection task, which quickly becomes inadequate when tenants need to configure not only functional features but also user‑interface layouts, security policies, deployment options, and other orthogonal concerns. The authors propose a Multi‑Dimensional Customization Model built on the concept of a metagraph—a hyper‑graph structure where edges can connect multiple input and output nodes—tailored to capture the richness and inter‑dependency of SaaS customizations.

First, the authors decompose customization into distinct dimensions: functional, UI, security, deployment, and potentially others. Each dimension is modeled as an independent sub‑metagraph that encodes its own elements (nodes) and relationships (edges) such as inclusion, exclusion, and dependency. For example, in the security dimension, “multi‑factor authentication” excludes “single‑factor authentication,” while in the UI dimension, “dark theme” may require the “high‑contrast mode” feature.

The core technical contribution is a composition algorithm that merges these sub‑metagraphs into a single, global metagraph representing the entire SaaS product line. During composition, common nodes (e.g., shared data models) and common edges (e.g., global compliance policies) are identified, and cross‑dimensional conflicts are resolved using a hierarchy of rules: priority‑based (security policies outrank UI preferences), policy‑based (corporate governance mandates), or tenant‑defined custom rules. The resulting global metagraph defines the feasible configuration space for all tenants.

To guarantee that tenant‑provided customizations are valid, the authors implement a metagraph‑based verification engine. A tenant’s selection is mapped onto the global metagraph, and three fundamental constraints are checked: existence (the selected element exists in the model), inclusion (all required predecessor elements are present), and exclusion (no mutually exclusive elements are simultaneously chosen). These checks are translated into logical formulas and fed to a SAT/SMT solver, enabling near‑real‑time feedback on configuration errors.

The experimental evaluation simulates a large SaaS platform with over 1,000 synthetic tenants, each configuring ten dimensions with an average of 20 options per dimension. Compared with a conventional single‑dimensional feature model, the metagraph approach shows only modest overhead: verification time grows from 45 ms to 80 ms even when the number of options increases tenfold, and conflict detection accuracy rises from 85 % to 99.3 %. Memory consumption scales linearly with the number of dimensions, confirming the model’s scalability.

The paper’s contributions can be summarized as follows: (1) a novel multi‑dimensional metagraph modeling framework that captures both intra‑ and inter‑dimensional customization relationships; (2) an algorithm for systematic composition of independent dimension models while automatically handling cross‑dimensional constraints; (3) a verification engine that leverages SAT/SMT solving to provide immediate, precise validation of tenant configurations; and (4) empirical evidence demonstrating that the approach remains efficient and accurate at scale.

In conclusion, the proposed multi‑dimensional metagraph model offers SaaS providers a practical pathway to deliver highly customizable services to a massive tenant base without sacrificing system integrity or incurring excessive engineering effort. By formalizing customization variability across orthogonal dimensions and providing automated composition and validation mechanisms, the approach promises reduced operational costs, improved tenant satisfaction, and stronger compliance guarantees. Future work suggested by the authors includes integrating machine‑learning‑driven recommendation engines to suggest optimal configurations and embedding the metagraph verification step directly into continuous deployment pipelines for seamless, zero‑downtime customization rollout.