Abstract | ||
---|---|---|
In recent years, while the Internet has brought various technological evolutions, a lot of ontology is required to organize and systemize knowledge, and its generation is necessary. Especially, classification of hypernym-hyponym relation which describes taxonomy of ontology has received a lot of attention. As a method to automate the generation, word embedding based method was proposed recently. Although the method enabled high accuracy classification by using semantics, it does not correspond to complex term consisting of multiple words. Based on this background, in this paper, we proposed a new model combined word embedding and Recurrent Neural Network(RNN), evaluated the classification performance with data extracted from WordNet. For the result, it is indicated that the RNN approach is more effective and general for ontology generation. |
Year | DOI | Venue |
---|---|---|
2019 | 10.1109/ICCC.2019.00024 | 2019 IEEE International Conference on Cognitive Computing (ICCC) |
Keywords | Field | DocType |
Ontological Classification,Word Embedding,Word2Vector,Recurrent Neural Network,Natural Language Processing,Recurrent Neural Network Language Model | Ontology,Computer science,Recurrent neural network,Natural language processing,Artificial intelligence,Word embedding,WordNet,Semantics,The Internet | Conference |
ISBN | Citations | PageRank |
978-1-7281-2712-5 | 0 | 0.34 |
References | Authors | |
0 | 2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Atsushi Oba | 1 | 0 | 0.34 |
Incheon Paik | 2 | 241 | 38.80 |