[Paper Review] Generating Multilingual Personalized Descriptions of Museum Exhibits - The M-PIRO Project

Ion Androutsopoulos, Vassiliki Kokkinaki+4 more|ArXiv.org|Oct 29, 2001

Natural Language Processing TechniquesComputer Science21 references22 citations

TL;DR

This paper presents the M-PIRO project, which develops a multilingual, personalized description generation system for museum exhibits using a language-independent database and user-specific profiles. The system dynamically generates tailored textual or spoken descriptions in English, Greek, and Italian based on user demographics and interaction history, with a web-based demonstrator validating its functionality in multilingual virtual environments.

ABSTRACT

This paper provides an overall presentation of the M-PIRO project. M-PIRO is developing technology that will allow museums to generate automatically textual or spoken descriptions of exhibits for collections available over the Web or in virtual reality environments. The descriptions are generated in several languages from information in a language-independent database and small fragments of text, and they can be tailored according to the backgrounds of the users, their ages, and their previous interaction with the system. An authoring tool allows museum curators to update the system's database and to control the language and content of the resulting descriptions. Although the project is still in progress, a Web-based demonstrator that supports English, Greek and Italian is already available, and it is used throughout the paper to highlight the capabilities of the emerging technology.

Motivation & Objective

• To develop a system that automatically generates multilingual, personalized descriptions of museum exhibits for web and virtual reality environments.
• To enable customization of descriptions based on user-specific factors such as age, background, and prior interaction with the system.
• To provide museum curators with an authoring tool for updating content and controlling language and style of generated descriptions.
• To support multilingual access to cultural heritage collections through dynamic text generation in multiple languages.
• To demonstrate the feasibility of the system with a working web-based prototype supporting English, Greek, and Italian.

Proposed method

• The system uses a language-independent database storing exhibit metadata and descriptive fragments.
• Natural language generation (NLG) techniques are applied to compose coherent, contextually appropriate descriptions from structured data and user profiles.
• User profiles—including age, background, and interaction history—are used to modulate content and linguistic style.
• Descriptions are generated in multiple languages (English, Greek, Italian) via language-specific templates and linguistic rules.
• An authoring tool allows curators to manage the knowledge base, define templates, and control language and content parameters.
• A web-based demonstrator integrates the system components and enables real-time generation of multilingual, personalized exhibit descriptions.

Experimental results

Research questions

• RQ1 How can multilingual, personalized descriptions of museum exhibits be automatically generated from a shared, language-independent knowledge base?
• RQ2 To what extent can user-specific factors such as age and background influence the content and style of generated descriptions?
• RQ3 What is the feasibility of integrating dynamic description generation into virtual museum environments accessible via the web?
• RQ4 How can museum curators effectively control and update the content and language of generated descriptions?
• RQ5 Can a prototype system successfully deliver personalized, multilingual exhibit descriptions in real time across multiple languages?

Key findings

• The M-PIRO system successfully generates multilingual, personalized descriptions of museum exhibits using a centralized, language-independent database.
• User-specific customization based on age, background, and prior interaction is effectively implemented and demonstrated in the web-based prototype.
• The system supports dynamic content generation in English, Greek, and Italian, with consistent linguistic quality across languages.
• The authoring tool enables curators to update content and control language and style, ensuring maintainability and flexibility.
• A functional web-based demonstrator is operational and validates the core architecture and capabilities of the system.
• The project demonstrates the feasibility of scalable, multilingual, personalized narrative generation for cultural heritage applications.