Graph Convolutional Network for Word Sense Disambiguation

Table 1

Disambiguation accuracies in the first group of experiments.

Ambiguous words	LSTM	CNN	CNN + LSTM	CNN + Bi-LSTM	GCN(1)	GCN(2)
Biaomian(2)	0.611	0.611	0.611	0.611	0.833	0.889
Cai(2)	0.524	0.550	0.476	0.524	0.737	0.737
Danwei(2)	0.632	0.632	0.632	0.632	0.882	0.706
Dongyao(2)	0.625	0.625	0.625	0.625	0.563	0.688
Ernv(2)	0.545	0.524	0.545	0.545	0.850	1.000
Jingtou(2)	0.533	0.533	0.533	0.533	0.568	0.600
Kaitong(2)	0.500	0.526	0.500	0.500	0.500	0.700
Qixi(2)	0.286	0.714	0.714	0.714	0.714	1.000
Qixiang(2)	0.647	0.647	0.647	0.647	0.688	0.938
Shi(2)	0.625	0.625	0.625	0.625	0.813	0.813
Tuifan(2)	0.400	0.600	0.600	0.600	0.600	0.800
Wang(2)	0.769	0.769	0.769	0.769	0.769	0.769
Yanguang(2)	0.714	0.714	0.714	0.714	0.571	0.714
Zhenjing(2)	0.714	0.714	0.714	0.714	0.857	0.786
Zhongyi(2)	0.480	0.458	0.480	0.480	0.813	0.688
Ben(3)	0.429	0.370	0.429	0.429	0.760	0.760
Bu(3)	0.500	0.522	0.500	0.500	0.750	0.850
Chengli(3)	0.357	0.370	0.357	0.357	0.357	0.393
Duiwu(3)	0.458	0.478	0.458	0.417	0.546	0.636
Gan(3)	0.450	0.421	0.450	0.450	0.889	0.611
Qizhi(3)	0.556	0.556	0.556	0.556	0.667	0.722
Rizi(3)	0.333	0.344	0.333	0.333	0.406	0.594
Tiandi(3)	0.400	0.375	0.400	0.400	0.800	0.720
Tiao(3)	0.438	0.438	0.438	0.438	0.643	0.714
Changcheng(3)	0.577	0.600	0.577	0.577	0.524	0.571
Chi(4)	0.500	0.259	0.500	0.500	0.478	0.565
Dong(4)	0.542	0.528	0.542	0.542	0.500	0.500
Jiao(4)	0.250	0.263	0.250	0.250	0.359	0.410
Average accuracy	0.514	0.527	0.535	0.535	0.659	0.710