Integrating Wiki Systems, Natural Language Processing, and Semantic Technologies for Cultural Heritage Data Management

Abstract

Modern documents can easily be structured and augmented to have the characteristics of a semantic knowledge base. Many older documents may also hold a trove of knowledge that would deserve to be organized as such a knowledge base. In this chapter, we show that modern semantic technologies offer the means to make these heritage documents accessible by transforming them into a semantic knowledge base. Using techniques from natural language processing and Semantic Computing, we automatically populate an ontology. Additionally, all content is made accessible in a user-friendly Wiki interface, combining original text with NLP-derived metadata and adding annotation capabilities for collaborative use. All these functions are combined into a single, cohesive system architecture that addresses the different requirements from end users, software engineering aspects, and knowledge discovery paradigms. The ideas were implemented and tested with a volume from the historic Encyclopedia of Architecture and a number of different user groups.

Natural language processing for MediaWiki

“…In contrast to Wikulu and our previous work in the Durm project [9], in this paper we offer a general architecture that allows any wiki to benefit from various NLP techniques brokered via the Semantic Assistants framework. In other words, our Wiki-NLP integration is not only NLP service independent, but also offers a flexible and easily extensible design for the integration of various wiki engines with NLP capabilities.…”

“…To preserve these artifacts, they are often digitized and stored in nationally and internationally distributed databases. We previously developed a custom wiki integration for this scenario, the Durm Wiki, 17 where we demonstrated how modern semantic technologies offer the means to make these heritage documents accessible by transforming them into a semantic knowledge base [9]. Now we are able to offer the same NLP capabilities that had been implemented in the Durm wiki using our generic Wiki-NLP integration architecture, in addition to all the other services made available through the general Semantic Assistants framework.…”

“…In our application scenario, a historical encyclopedia of architecture, this index helped domain experts to locate information that they could not find through information retrieval methods: since the 100-year old heritage documents contain outdated terminology no longer in use, full-text search was not sufficient to discover concepts [9]. Additional services provided summaries of wiki content based on a specific question [9]. These ideas can be applied to other heritage data projects and knowledge management wikis in general.…”

“…The presence of an index page in the wiki that is automatically maintained not only aggregates information on a high level and helps users to find information 'at a glance', but also enables them to "discover" interesting concepts or entities that they did not know were present in the wiki. In our application scenario, a historical encyclopedia of architecture, this index helped domain experts to locate information that they could not find through information retrieval methods: since the 100-year old heritage documents contain outdated terminology no longer in use, full-text search was not sufficient to discover concepts [9]. Additional services provided summaries of wiki content based on a specific question [9].…”

Supporting wiki users with natural language processing

“…The generated index page, as shown in Figure 4, presents an alphabetically-ordered list of wiki terms and a direct link to their pages inside the wiki. In experiments with end users, we found that the presence of such an automatically maintained index page not only aggregates the wiki's embodied content on a high-level and enables users to find information at a glance, but also helps them to "discover" interesting concepts or entities that they did not know were present in the wiki [6]. platform, have demonstrated their capabilities in requirements engineering processes.…”