Event argument extraction (EAE) is a sub-task of event extraction that identifies and extracts relevant information or arguments from events. In natural language processing and information extraction tasks, an event usually comprises an action performed by one or more entities at a certain time and location. The participants and arguments involved in the event are identified and their roles are determined in EAE. In the current state-of-the-art, prompt-based approaches are used to ask pre-trained language models to extract arguments about input context. Existing prompt-based techniques, however, hinge on separated, constructed by hand prompts that fail to take context associated with specific event types into account to enrich prompts for EAE. Moreover, the recently proposed prompt learning models require a great deal of template annotation. Therefore, we suggest a new prompt-based technique for Sentence-level EAE (SEAE) which presents soft prompts to promote the aggregation of specific event types and related input sentences context to improve EAE. Method we suggestd in this paper is extensively evaluated on a sentence-level benchmark dataset for EAE. The results present promising improvements that we achieves a substantial improvement by comparing method we propose to previous ones. Furthermore, further analysis confirms its effectiveness and generalization to few-shot settings.