Open maxkratz opened 3 years ago
Another idea to improve the embedding quality as well as the performance: Rating of matches:
Ideas:
Example criteria:
The cost function (e.g., with input = all attributes of all nodes of a match) could, for example, be learned by ML approaches.
Another idea to improve the embedding quality as well as the performance: Rating of matches:
Ideas:
Example criteria:
The cost function (e.g., with input = all attributes of all nodes of a match) could, for example, be learned by ML approaches.