Model-based information navigation for engineering documents

David Jones,Chris Snider,Jason Matthews,Jason Yon,Jeff Barrie,Kevin Robinson,Ben Hicks

doi:10.1016/j.compind.2020.103254

Abstract

Engineering and the manner in which engineers think is largely visual and functional, and yet engineers are typically provided with search engines that are text-based. While software based on a visual and functional ethos exist (CAD for example), when searching for information engineers are still required to enter a text query into a search box. This process holds potential incompatibilities both with the nature of the data (i.e. 3D models) and with the way in which engineers think and work. Consequentially, the proposition tested in this paper is that a model-based approach to information access, i.e. a representation of an organisations information around a model of an artefact i.e. CAD model, can improve engineering information retrieval.In an A-B test with a traditional text-based search engine, and using study questions derived from real-world information seeking scenarios based on the activities of a world-leading aircraft manufacturer, the results presented in this paper suggest that there is merit to such an approach. Specifically, this paper shows that there is no significant difference in time to complete a search between a model-based and text-based interfaces in spite of the addition of a new stage in the search process (navigating a 3D model); that the system structure of the model-based interface allows for non-text based documents to be indexed, making up for the inherent limitations in traditional text-based search; and that participants enjoy using the model-based interface and find it intuitive, easy and simple to use. Further, this paper also finds that those with more experience/familiar with the product structure and those in managerial positions are more likely to find information using a model-based interface that those who are not, who perform better using a text-based interface.

Highlights

Knowledge, information, and data are widely considered to be an organisation’s greatest asset and as such, the effective capture, use and re-use of this asset is deemed as good practice [1]
A model-based approach to information retrieval transcends a number of interrelated fields such as information search, engineering information search, model-based approaches to engineering, and model-based information navigation
Work in the area of improving Intranet/enterprise search is on-going one area of the problem space that has not received much academic consideration is the form of the interface and opportunities to create interfaces that capitalise on the visual and functional nature of how engineers think and the engineering process. This paper addresses this gap by building on model-based approaches to engineering and developing and evaluating a modelbased information navigation system

Summary

Introduction

Information, and data are widely considered to be an organisation’s greatest asset and as such, the effective capture, use and re-use of this asset is deemed as good practice [1]. Internet search engines touch most our lives with the major search providers of Google and Microsoft seemingly becoming the doorways into humanity’s data, information, and knowledge. There is a disconnect between the advances of Internet search technology and the knowledge management needs of organisations and this is true for engineering organisations [10]. It is to this field that this paper contributes. At the point when a user enters a search query, the search engine does not trawl through every document in the corpus, that would be highly inefficient and take a long time. The corpus is preprocessed, and stored as an index, with the inverted index being one of the most common and efficient types [16]

Objectives

Methods

Results

Discussion

Conclusion