Semantic modeling and natural language processing for environmental compliance checking

Abstract

There is an increasing need for improved environmental compliance, while reducing the time and cost of compliance checking (CC). This research offers a new approach for automated environmental CC. The approach utilizes semantic modeling and natural language processing (NLP) techniques. Semantic modeling aims at offering the level of knowledge representation and reasoning that is needed to process applicable environmental regulations and check compliance of construction plans to the rules that are prescribed by those regulations. NLP techniques will facilitate text analysis and processing for achieving human-like extraction and formalization of rules and information. The approach is intended to automatically detect non-compliance instances and provide a rich analysis of the non-compliance such as which regulation was violated, reason for violation, possible consequences of violation. To achieve such deep levels of text processing and automated reasoning, three algorithms are developed and combined into one computational platform: (1) a machine-learning-based text classification algorithm to classify relevant text (in documents such as environmental regulations), (2) a hybrid syntactic-semantic, utilizing grammatical and meaning-descriptive features of the text, information extraction algorithm to facilitate text processing for extraction and formalization of rules and information, and (3) a logic-based algorithm for compliance reasoning. Automated analysis will be facilitated by a semantic model for environmental knowledge representation and reasoning. The algorithms will be implemented in a proof-of-concept prototype software for automated environmental CC, and will be tested and validated using real-life-based test case scenarios. This research will transform the way we conceptualize and formally reason about complex regulatory schemes and associated CC processes, and will advance the research in the areas of deep NLP and deep semantic reasoning. The results of this research will also transform the way construction professionals and government regulators (e.g. environmental protection agencies) check the compliance of construction projects with environmental regulations and green practices. The ultimate goals of this research, in terms of benefits to the society, are: (1) increasing environmental compliance in construction, thereby protecting human health and the environment, (2) promoting compliance with non-regulatory green construction practices, thereby supporting green and sustainable construction, and (3) reducing the time and cost of environmental CC.2 2 This material is based upon work supported by the National Science Foundation under Grant No. 1201170. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author and do not necessarily reflect the views of the National Science Foundation.