The aim of this study is to identify the flow structure in an airlift pump–bubble generator system by using experimentally obtained differential pressure signals. Differential pressure measurements were applied at both the bottom and top test sections. The normalized time variation of the differential pressure data was analyzed using the probability density function (PDF), power spectral density function (PSDF), Kolmogorov entropy, and discrete wavelet transform (DWT). The results indicate that the water movement mechanism in the riser pipe could be divided into three regions, namely fixed liquid, locally moving liquid, and fully moving liquid, depending on the supplied superficial air velocity. Moreover, the chaotic in riser pipe increases with the increase of the submergence ratio at high supplied superficial air velocity. The chaotic level at the bottom of the test section was higher than that at the top of the test section at low supplied superficial air velocity. Finally, the observed flow patterns closely conformed to the previous definition and flow pattern maps. The bubbly flow was classified into clustered, homogeneous, and cap bubble regimes. The slug flow regime was classified into bubbly stable slug, bubbly unstable slug, and slug-churn, wherein the frequencies of both the liquid and air slugs increased with an increase in the air superficial velocity and submergence ratio.