This paper introduces a novel approach for outlier detection in high-dimensional data called Angle-Based Outlier Detection (ABOD). ABOD uses the variance in angles between difference vectors of data points to identify outliers, which is less sensitive to the "curse of dimensionality" compared to purely distance-based methods. ABOD does not require parameter selection, making it more efficient and robust. The paper compares ABOD with the well-established LOF method on various artificial and real-world datasets, showing that ABOD performs particularly well in high-dimensional data. ABOD is based on the idea that outliers have a different distribution of angles compared to typical data points. The paper also presents variants of ABOD, including FastABOD for large low-dimensional datasets and LB-ABOD for high-dimensional datasets. The results show that ABOD provides better precision and recall in outlier ranking, especially in high-dimensional data. The paper also discusses the efficiency of ABOD and its variants, showing that LB-ABOD significantly improves performance by reducing computation time. The paper concludes that ABOD is a promising method for outlier detection in high-dimensional data.This paper introduces a novel approach for outlier detection in high-dimensional data called Angle-Based Outlier Detection (ABOD). ABOD uses the variance in angles between difference vectors of data points to identify outliers, which is less sensitive to the "curse of dimensionality" compared to purely distance-based methods. ABOD does not require parameter selection, making it more efficient and robust. The paper compares ABOD with the well-established LOF method on various artificial and real-world datasets, showing that ABOD performs particularly well in high-dimensional data. ABOD is based on the idea that outliers have a different distribution of angles compared to typical data points. The paper also presents variants of ABOD, including FastABOD for large low-dimensional datasets and LB-ABOD for high-dimensional datasets. The results show that ABOD provides better precision and recall in outlier ranking, especially in high-dimensional data. The paper also discusses the efficiency of ABOD and its variants, showing that LB-ABOD significantly improves performance by reducing computation time. The paper concludes that ABOD is a promising method for outlier detection in high-dimensional data.