Initial Development of the Engineering Genome Project--an Engineering Ontology with Multimedia Resources for Teaching and Learning Engineering Mechanics

This paper reports on the development of an engineering mechanics ontology in support of the Engineering Genome Project (EGP). Ontologies contain both hierarchical (i.e., taxonomic) and relationship information, and they organize knowledge into highly-structured and expertly-constructed schemas. Ontologies can help facilitate the novice-to-expert transition for learners by making the implicit connections and relationships among complicated and cross-linked pieces of information (that experts know but novices do not) explicit. Our goal with the EGP is to empower learners to explore these detailed relationships among complicated concepts and topics in order to mediate the transition from novice to expert. This paper covers the initial development of the Engineering Genome itself – specifically, the development of the: (i) knowledge architecture and the classes in the ontology; (ii) the coarse-scale and fine-scale attributes used to describe the rich characteristics of the knowledge; and (iii) the relationship information describing the invisible-to-novices connections among disparate pieces of knowledge. We discuss key decisions we have made about the architecture of the knowledge, as well as the computer implementation of the ontology using the Web Ontology Language (OWL). The paper will follow this structure. First we introduce the notion of controlled vocabularies and the overall conceptualization of the knowledge domain. Next, we illustrate the development of specific relationships and attributes, and highlight the challenges of establishing a knowledge architecture for seemingly straight-forward concepts (such as coordinate systems and units). Next we show how the ontology can also be linked to specific curricula and in particular learning outcomes associated with courses (as well as ABET) in which specific concepts are introduced. Finally, we explain and demonstrate the query procedures through which the ontology is mined for relationship information that–despite their expertise–experts may not fully be aware of. Our results so far indicate that an ontology can indeed be developed for engineering mechanics, and that the potential pedagogical uses for a carefully-constructed knowledge architecture are promising. Introduction “Genomics” is a term that has entered the common English lexicon in recent years, and its definition has evolved from its strictly scientific origins. Genomics originally referred to the study of the structure, function, and organization of the chromosomes of an organism, but has more recently taken on the meaning of studying the (relationships among) underlying building blocks of a system. For instance, we now speak of organizational DNA, which expresses the underlying strength and interactions of an organization’s decision processes, information sharing, rewards systems, and management structures. Enterprise solutions for knowledge management present approaches to capture and archive an organization’s institutional knowledge from its systems, employees, and partners, and to operationalize that knowledge in the organization’s everyday practice (Shahnaqaz et al.1, and the explosion of knowledge management journals in the past 10 years2). But a more accessible example of genomic thinking comes from pop culture: the Music Genome Project3. About ten years ago, a group of music performers, experts, and enthusiasts came together with the goal of creating “the most comprehensive analysis of music ever.” They defined hundreds of P ge 23753.2 musical attributes (’genes’) that contain the granular, essential information about a particular piece of music. They then set about categorizing individual songs according to the taxonomy they had developed. So the Music Genome is a collection of digital assets (the songs) tagged with highly granular, descriptive attributes (the genes), and organized into a searchable database. The result is Pandora–the Internet radio station that allows users to probe the Music Genome and create playlists based upon keyword searches. Pandora interrogates the Music Genome to create a playlist of songs that are genomically related – the songs are close neighbors in the genome. Songs are not arranged by genre, or by era, or by band geographic origin, or by sales, or by any other coarse metric or hierarchical relationship. Rather, songs in the playlist are presented based upon their genomic similarity, based upon shared attributes. The power of the Music Genome is that it exposes relationships among songs that, on the surface, appear to be unrelated (or perhaps only tangentially related). The listener instantly becomes educated about, and curious to explore, these relationships. Now imagine that we create a similar structure not for music, but for the whole of engineering knowledge. This Engineering Genome would include multimedia learning objects, tagged with appropriate attributes and organized into a searchable database, and it would allow learners to interrogate its contents and explore the underlying relationships among the individual bits of engineering knowledge. When users interrogate the engineering genome, they will be presented with multimedia learning objects that are not organized by hierarchy, but rather based upon genomic similarity (along the lines of how Pandora matches songs genomically4. The EGP intentionally bridges traditional content silos (i.e., classes in the curriculum) and enables the learner to visualize and comprehend the elusive and subtle relationships among engineering concepts that, on the surface, appear to be unrelated (or perhaps only tangentially related). The Pandora metaphor is used in Table 1 to present a high-level view of the EGP. The EGP defines a multimedia learning object as a sharable digital file containing information useful for teaching and learning, and encompasses a wide range of file types (examples: video and audio files in Quicktime or MP4, Matlab .m files, java applets, PDF files, etc.). The specifics of each multimedia asset depend upon context; we currently have files spanning a range of different teaching and learning tools, including: (i) lecture videos, (ii) video problem solutions, (iii) simulations/animations, (iv) Matlab .m files and other executables, (v) text-based resources in PDF. Many others are possible and the EGP can admit these and many other file types. Learning “content” is, however, not enough; we want students to understand the relationships among seemingly disparate pieces of content. Since at least the 1970’s, there have been persistent calls5;6;7 for increasing synthesis and design in the engineering curriculum, for greater emphasis on deep inquiry, and for a general reversal of the compartmentalization of engineering content in the classroom. The ability to integrate knowledge is a key trait of the modern engineer, and traditional engineering curricula often struggle to instill this trait. The Engineering Genome, once fully realized, will address this critical need by building a cross-curricular tool that describes the incredible richness of relationships between pieces of “content”, and therefore promotes student understanding and integration of knowledge. Methods–Controlled Vocabularies P ge 23753.3 feature Pandora, The Music Genome The Engineering Genome the digital assets (the “content”) individual songs multimedia learning objects covering specific subjects in very detailed ways the “genes” rhythm, instrumentation, harmony, orchestration, lyrics, etc. (more than 400 in all) mathematical model, solution technique, perceived difficulty, etc. hierarchical relationships very limited: band → album → song more formal: branches and subbranches of engineering knowledge, along with taxonomies of mathematics and other disciplines TABLE 1. A high-level view of the EGP, using Pandora as a metaphor. The core of the Engineering Genome ontology is the underlying taxonomy itself, which is informed by supporting taxonomies (math, physics) and the discipline taxonomy of engineering mechanics, and contains a rich set of categories and tags in a relational database. The hierarchical organization of the supporting taxonomies presents itself as a parent-child relationship with set of coarse-scale attributes for all topical areas with the fine-scale taxonomy varying as appropriate by discipline. A controlled vocabulary, a set of standard terms, will be used for each of the supporting and discipline taxonomies. In traditional information systems (i.e., library catalog), controlled vocabularies help to bring together under a single word or phrase, all the material that is available on a particular topic. The main purpose is to provide some mechanism for querying multiple resources simultaneously and provide some commonality of description across the resources being made available for searching. Controlled vocabularies group similar objects together and ensure consistency for searching. The use of a predefined, authorized set of terms applied to objects by a domain expert improves the relevancy of a search result. Using terms from established taxonomies will allow the Genome to integrate with other systems or content (for example books and other items from a library catalog) if so desired. The Knowledge Domain and Its Description The Engineering Genome is both an expertly-constructed organization of engineering knowledge, as well as a library of multimedia learning objects that are categorized according to that organizational structure. As such, it takes an applied ontological view of the knowledge domain by emphasizing both hierarchical constructs (taxonomies) and relational features that do not necessarily follow parent-child lines. These relational features really do demand an ontological view rather than a strictly hierarchical one, and we will consistently use the term ontology to describe the Engineering Genome. For the purposes of this research project, an ontology is composed of several parts: (i) a structure