ITcon Vol. 13, pg. 193-211,

ABSTRACTOR: An agglomerative approach to interpreting building monitoring data

submitted:November 2006
revised:February 2008
published:May 2008
editor(s):Amor R
authors:Apkar Salatian
School of Computing, The Robert Gordon University, St Andrew Street, Aberdeen, AB25 1HG,
United Kingdom

Bruce Taylor
Scott Sutherland School, The Robert Gordon University, Garthdee Road, Aberdeen AB10 7QB,
United Kingdom
summary:Building operators are confronted with large volumes of continuous data from multiple environmental sensors which require interpretation. The ABSTRACTOR system under development summarises historical data for interpretation and building performance assessment. The ABSTRACTOR algorithm converts time series data into a set of linear trends which achieves data compression and facilitates the identification of significant events on concurrent data streams. It uses a temporal expert system based on associational reasoning and applies three consecutive processes: filtering, which is used to remove noise; interval identification to generate temporal intervals from the filtered data - intervals which are characterised by a common direction of change (i.e increasing, decreasing or steady); and interpretation which performs summarisation and assists building performance assessments. Using the temporal intervals, interpretation involves differentiating between events which are environmentally insignificant and events which are environmentally significant. Inherent in this process are rules to represent these events. These rules support temporal reasoning and encapsulate knowledge to differentiate between events.
keywords:Building, monitoring, interval identification, data mining, clustering, data compression
full text: (PDF file, 0.365 MB)
citation:Salatian A, Taylor B (2008). ABSTRACTOR: An agglomerative approach to interpreting building monitoring data, ITcon Vol. 13, pg. 193-211,