A Review on the Application of Knowledge Graph Technology in the Medical Field

Table 4

Classification table of entity linking methods.

Name	Rule	Advantages	Disadvantages
Entity attributes-based	Computing the similarity between the attributes	Simple realization and the high accuracy rate when the medical attribute is rich	With poor antinoise ability, the accuracy cannot be guaranteed when the entity attribute is sparse.
Entity popularity-based	Based on probability statistics and the frequency in the medical encyclopedia	Reliable and simple heuristic rules	Poor robustness and the ambiguity of specific entities are not considered.
Context-based	Based on the similarity between the entity contexts	High accuracy when the text is long enough and relatively clean	With less flexibility, the accuracy rate cannot be guaranteed when the text is sparse.
External evidence-based	Based on the semantic correlation	Introduced strong expansibility of rich feature information	Method effectiveness depends on the quality of external evidence.