(item-based) Pearson similarity

Adjusted cosine similarity (distance) - similarity that accounts for different baselines in dimensions $CosSim=\frac{{\sum }_{i=1}^n(x_i-\bar{l})(y_i-\bar{l})}{\sqrt{{\sum }_{i=1}^n(x_i-\bar{l}{)}^2}\sqrt{{\sum }_{i=1}^n(y_i-\bar{l}{)}^2}}$ $\bar{l}$ - average value of all attribute values