An Intuitive Interface for Technical Documentation Based on Semantic Knowledge Graphs

Abstract

Maintaining technical documentation is a challenge. Products are becoming more complex, product lifecycles are getting shorter, and the number of product variants is increasing. Manuals that guide personnel in the use and maintenance of products are critical to their efficient and safe operation. Authoring system for technical documentation therefore increasingly apply semantic models to control the cost of maintaining technical documentation. Working with formal semantic structures is challenging for technical writers who usually work with plain, written text. This paper presents an intuitive interface for a semantic knowledge graph to facilitate the adoption and use of semantic models in technical documentation. The interface allows working with unstructured text and to postpone its semantification. Users can add semantic annotations in an iterative and incremental way. The interface was developed using a user-centered design process and subjected to an evaluation with technical writers. The results indicate that technical writers could use the prototype successfully and enjoyed the underlying concepts. Further iterations will extend the system and, for example, use artificial intelligence to suggest semantic links to improve the quality of the knowledge graph.