图书与情报
圖書與情報
도서여정보
LIBRARY AND INFORMATION
2014年
5期
20-25
,共6页
大数据%术语自动抽取%关联规则
大數據%術語自動抽取%關聯規則
대수거%술어자동추취%관련규칙
big data%automatic term extraction%association rules
文章在文献调研的基础上,通过理论与实验结合的方法讨论了基于关联规则的术语抽取方法的合理性和可用性。从理论上看,关联规则的基本原理决定了它在充分解决“序”的条件下,可以解决术语的识别和抽取问题;从实践上看,关联规则的方法的确可以正确抽取出术语,而且,通过与现有算法的比较,可以发现,关联规则在算法实现难度和算法占用资源方面具有较明显的优势。
文章在文獻調研的基礎上,通過理論與實驗結閤的方法討論瞭基于關聯規則的術語抽取方法的閤理性和可用性。從理論上看,關聯規則的基本原理決定瞭它在充分解決“序”的條件下,可以解決術語的識彆和抽取問題;從實踐上看,關聯規則的方法的確可以正確抽取齣術語,而且,通過與現有算法的比較,可以髮現,關聯規則在算法實現難度和算法佔用資源方麵具有較明顯的優勢。
문장재문헌조연적기출상,통과이론여실험결합적방법토론료기우관련규칙적술어추취방법적합이성화가용성。종이론상간,관련규칙적기본원리결정료타재충분해결“서”적조건하,가이해결술어적식별화추취문제;종실천상간,관련규칙적방법적학가이정학추취출술어,이차,통과여현유산법적비교,가이발현,관련규칙재산법실현난도화산법점용자원방면구유교명현적우세。
On the basis of sufficient literature review, the rationality and availability of automatic term extraction based on association rules are discussed by the theoretical and experimental methods. Theoretically, the basic principle of association rule, under the condition of full solution of the "sequence", can solve the problem of identification and extraction of the term. Practically, association rules method can extract correct terminology, and by comparing with the existing algorithm, association rules algorithm has more obvious advantages in difficulty of realization and occupied resources.