Spike-timing-dependent plasticity (STDP) is a biological process in which the precise order and timing of neuronal spikes affect the degree of synaptic modification. While there has been numerous research focusing on the role of STDP in neural coding, the functional implications of STDP at the macroscopic level in the brain have not been fully explored yet. In this work, we propose that STDP in an ensemble of spiking neurons renders storing high dimensional information in the form of a `memory plane'. Neural activity based on STDP transforms periodic spatio-temporal input patterns into the corresponding memory plane, where the stored information can be dynamically revived with a proper cue. Using the dynamical systems theory that shows the analytic relation between the input and the memory plane, we were able to demonstrate a specific memory process for high-dimensional associative data sets. In the auto-associative memory task, a group of images that were continuously streamed to the system can be retrieved from the oscillating neural state. The second application deals with the process of semantic memory components that are embedded from sentences. The results show that words can recall multiple sentences simultaneously or one exclusively, depending on their grammatical relations. This implies that the proposed framework is apt to process multiple groups of associative memories with a composite structure.