From Text to Databases: attribute grammar as database meta-model

TL;DR

The paper tackles the challenge of turning unstructured textual data into structured database representations by evolving both the textual instance and a meta-model $\mathbb{G}$ defined as an attribute grammar. Starting with an initial grammar $G_0$, the method iteratively derives a target grammar $G_T$ that satisfies $\mathbb{G}$ through tree enrichment, equivalence-class computation, and rewriting guided by similarity measures. A proof-of-concept using the CAS French Corpus demonstrates that the approach can produce a coherent, model-agnostic database schema $G_T$ and its corresponding instance, with potential mappings to relational or graph models. The work advances automated database structuring from text and highlights the balance between semantic enrichment and syntactic regrouping, enabling practical deployment in domains with rich textual sources such as clinical case descriptions.

Abstract

We present a general methodology for structuring textual data, represented as syntax trees enriched with semantic information, guided by a meta-model G defined as an attribute grammar. The method involves an evolution process where both the instance and its grammar evolve, with instance transformations guided by rewriting rules and a similarity measure. Each new instance generates a corresponding grammar, culminating in a target grammar GT that satisfies G. This methodology is applied to build a database populated from textual data. The process generates both a database schema and its instance, independent of specific database models. We demonstrate the approach using clinical medical cases, where trees represent database instances and grammars act as database schemas. Key contributions include the proposal of a general attribute grammar G, a formalization of grammar evolution, and a proof-of-concept implementation for database structuring.

Figures (18)

• Figure 1: An example of a graph database instance generated by structuring a text describing a clinical case.
• Figure 2: Iterative process for automatic structuring
• Figure 3: Example of entity incorporation in a syntax tree and simplifications
• Figure 4: Example of quotient tree
• Figure 5: An example of a tree instance after our iterative process, where internal nodes represent concepts from $\mathbb{G}$, along with its corresponding grammar $G_T$.

Theorems & Definitions (29)

• Example 1
• Definition 1: Ordered tree
• Definition 2: Sub-tree
• Example 2
• Definition 3
• Definition 4: Parse Tree
• Example 3
• Definition 5: Attribute Grammar
• Example 4
• Definition 6: Meta-grammar