This paper aims at developing a methodology that helps in the characterization of the effects of a single point and multipoint sparger on the flow patterns in a bubble column reactor. The methodology uses the simultaneously measured instantaneous velocity-time data (two orthogonal velocity components) using LDA at different measurement locations. The data sets are transformed into their equivalent description in terms of polar coordinates, viz., resultant velocity vector and angle. This information is further transformed in terms of symbolic representation in a binary number format. The covariance matrices obtained from the binary data are analyzed for their eigenvalues, and the spectrum is used for the identification of the dominant, energy containing modes in the data. In the vicinity of the sparger, the nature of covariance matrix and the eigenvalue spectrum for the single point sparger (SPS) data was significantly different from that of the multipoint sparger (MPS). In the bulk region, for both types of spargers, the nature of the eigenvalues and the number of dominant modes is seen to be different from the near sparger region. In general, more numbers of eigenvalues are required to capture the total energy in the data for a multipoint sparger when compared to a single point sparger. This suggests a higher level of complexities existing due to the intricate hydrodynamics for the multipoint sparger. The possible use of this methodology for practical purposes, e.g., identification of malfunctioning of gas-liquid sparger reactors is discussed.