Bibliometric and text mining approaches to evaluate landfill design standards

Abstract

In 2014, Canadians generated 961 kg of waste per capita. Landfilling is a logical choice for many Canadian communities because of land availability. This paper examines and compares five different design criteria from provincial standards and guidelines in British Columbia, Alberta, Manitoba, Nova Scotia, and the Northwest Territories. Text mining including word counts, word frequency analysis, lexical quantification, and the Gunning-Fog Index are used to identify linguistic and stylistic features in the corpora. Results show that design standards and guidelines tend to be driven by climate, demographic, and environmental considerations. Rank–frequency analysis showed that the design standards were non-Zipfian (0.409 < exponent < 0.681), but a power relationship generally fit well (0.90 < R2 < 0.99). Lexical density and the Gunning-Fog Index do not consider content in figures and tables, which are key components in engineering design standards and guidelines, so a Table and Figure Factor was developed. The factor was negatively (R2 = 0.735; Slope = − 1.1) related to lexical density, supporting the use of tables and figures in engineering design standards and guidelines. Comparing equivalent word counts in standards to landfilling rate and landfills per capita yielded better results (Spearman R = − 0.7, p = 0.08) compared to government expenditure on waste management. Lexical density and Gunning-Fog Index may be good metrics when comparing landfill efficiency metrics. The study’s outcomes may help to develop better landfill design standards in English speaking countries.