Natural language processing methods for knowledge management—Applying document clustering for fast search and grouping of engineering documents

Product development companies collect data in form of Engineering Change Requests for logged design issues, tests, and product iterations. These documents are rich in unstructured data (e.g. free text). Previous research affirms that product developers find that current IT systems lack capabilities to accurately retrieve relevant documents with unstructured data. In this research, we demonstrate a method using Natural Language Processing and document clustering algorithms to find structurally or contextually related documents from databases containing Engineering Change Request documents. The aim is to radically decrease the time needed to effectively search for related engineering documents, organize search results, and create labeled clusters from these documents by utilizing Natural Language Processing algorithms. A domain knowledge expert at the case company evaluated the results and confirmed that the algorithms we applied managed to find relevant document clusters given the queries tested.

TRANSLATION PECULIARITIES IN OIL AND GAS INDUSTRY

The article focuses on the actual issues of translation problems in the oil and gas industry, in particular on finding adequate equivalents to highly technical terms in both languages, the translation of abbreviations and the application of technical normative documentation (GOST[1]). The research is based on the engineering documents developed in English and Russian for the major oil and gas projects to be implemented in Russia 2006-2009.[1] Russian National Standard

DEPLOYING ENGINEERING CASES TO FACILITATE PROBLEM-BASED THINKING IN ENGINEERING COMMUNICATIONS

First-year engineering students often struggle to communicate the value of their work because they do not understand how problem-based reasoning drives engineering research and industry. Recognizing the effectiveness of discipline-specific teaching of the conventions of engineering communications, researchers have recently suggested the value of teaching the Swales "CARS" model to help students contextualize and justify their work. In two sections of Communications for the Engineering Profession at the University of Waterloo, we incorporated teaching of the Swales model of problem-based reasoning to help students understand the conventions of engineering communications, but found that authentic engineering documents are often too complex for this purpose. To address this limitation, we deployed engineering cases in two electrical/computer engineering courses to exemplify this model, and used pre-teaching and post-teaching surveys to measure students' perceptions of improvement in their ability to understand problem-based reasoning and apply it to project conceptualization. The results show that using simplified engineering cases of this kind both improves students' ability to use this model and improves their confidence in doing so. This outcome has implications for increasingly popular engineering-communications courses because it demonstrates the value of using realistic but simplified engineering scenarios to teach the Swales model in authentic and effective engineering communication.

Supporting Knowledge Re-Use with Effective Searches of Related Engineering Documents - A Comparison of Search Engine and Natural Language Processing-Based Algorithms

Product development companies are collecting data in form of Engineering Change Requests for logged design issues and Design Guidelines to accumulate best practices. These documents are rich in unstructured data (e.g., free text) and previous research has pointed out that product developers find current it systems lacking capabilities to accurately retrieve relevant documents with unstructured data. In this research we compare the performance of Search Engine & Natural Language Processing algorithms in order to find fast related documents from two databases with Engineering Change Request and Design Guideline documents. The aim is to turn hours of manual documents searching into seconds by utilizing such algorithms to effectively search for related engineering documents and rank them in order of significance. Domain knowledge experts evaluated the results and it shows that the models applied managed to find relevant documents with up to 90% accuracy of the cases tested. But accuracy varies based on selected algorithm and length of query.