FGNET-RH: Fine-Grained Named Entity Typing via Refinement in Hyperbolic Space

Zero-shot fine-grained entity typing in information security based on ontology

“…Several studies on entity typing have been conducted in the past [9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25]. In general, these studies can be divided into unsupervised and semi-supervised groups depending on the approach used.…”

“…Generally, such research focuses on the general field [9][10][11][12][13][14][15][16][17][18][19][20], and to the best of our knowledge, no such studies exist in the information security domain. The primary reasons are as follows.…”

“…(i) Although the general domain possesses certain mature knowledge bases that can be employed for fine-grained entity typing, such as Wikipedia, WordNet, and DBpedia, but they are not professional enough for the field of information security. Thus, studies such as [9][10][11][12][13][14][15], are not suitable for application in this domain. (ii) Furthermore, there is the absence of scaled, labelled data in the information security domain for use as a training corpus for classification models.…”

“…In addition, from the related literature on fine-grained entity typing [9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25], we determined that few researchers have considered the problem of misclassification in coarse-grained entity typing tasks. However, the inaccuracy in the results of coarse-grained typing certainly results in performance degradation in fine-grained entity typing tasks.…”