Repetition establishes a relationship with the listeners that is emotionally satisfying to our brains, and breeds a pleasure and memorable listening experience. In this paper, we present a soundscape linkage databank system which assists listeners to select music and lead to a chain of music clip with similar emotion. In the condition of choosing the melody by emotion, the autocorrelation function is applied for analyzing the period of the repeating musical structure which includes the Intro, Verse, and the Chorus. Then using this repeating musical structure to analyze the ingredients of music emotion by feature-weighted scoring algorithms and calculate the similarity of emotion between any two segments of the repeating musical structure to rank the resembling music. The study also demarcates the boundaries of four emotions in the emotion plane by Gaussian mixture model, and draws the locus of emotions evoked by musical signals. Preliminary evaluations indicate that the proposed algorithms produce results agreed approximately.