For Chinese Traditional Music (CTM) Knowledge Base: the accuracy of extracting entities from the CTM's ontology is insufficient

[candlecao]

This issue is subsequent to #21 (Subgraph Extraction)

As of 2024Nov27, I have basically finished "Agents Development" Hand-Over with my collaborator in China.
The Agents Development is for NLQ2SPARQL against CTM.

We observed the accuracy of entities (classes and properties) extraction from the CTM ontology is insufficient, which subsequently affects the precision of Subgraph extraction and, consequently, the effectiveness of NLQ2SPARQL.

To obtain better NLQ2SPARQL result, the prompts in the process should be refined.