deletions | additions       diff --git a/ComputationChordExtract.tex b/ComputationChordExtract.tex  index 95bffa6..4d4b62e 100644  --- a/ComputationChordExtract.tex  +++ b/ComputationChordExtract.tex 
...
\subsection{The Divergence}  In chapter~\ref{chintro} notions of theory, human subjectivity, and analytical technique were shown to govern classification of information from musical songs; MIR attempts to reconcile the differences between these approaches in an automated fashion, but ground-truth data is difficult to find or even psychologically establish. justify.  This paper focuses on how analysis can be performed on music informatics retrieval (MIR) tasks using computationally analyzed data compared with human-derived ground-truth data, with a focus on chord progressions. Subjective perception is emphasized important  as music only exists through the ears of a listener who often is informed and subconsciously guided by cultural cues, but cues;  analytical computation is and statistical methods  also exhibit strong explanatory power and are  shown to have great explanatory power. yield novel insights.  % "chord labels form an abstraction of musical content with substantial explanatory power" \cite{de2008tonal}.  \subsection{Harmony}  One question  \subsection{Background}  Chord progressions