《電子技術(shù)應(yīng)用》
您所在的位置:首頁(yè) > 人工智能 > 設(shè)計(jì)應(yīng)用 > 具有關(guān)系敏感嵌入的知識(shí)庫(kù)錯(cuò)誤檢測(cè)
具有關(guān)系敏感嵌入的知識(shí)庫(kù)錯(cuò)誤檢測(cè)
2020年信息技術(shù)與網(wǎng)絡(luò)安全第10期
繆 琦,楊昕悅
遼寧工程技術(shù)大學(xué) 電子與信息工程學(xué)院,遼寧 葫蘆島125105
摘要: 準(zhǔn)確性與質(zhì)量對(duì)于知識(shí)庫(kù)而言尤為重要,,盡管已經(jīng)有很多關(guān)于知識(shí)庫(kù)不完整性的研究,,但是很少有工作者考慮到對(duì)于知識(shí)庫(kù)存在的錯(cuò)誤進(jìn)行檢測(cè),按照傳統(tǒng)方法通常無(wú)法有效捕捉知識(shí)庫(kù)中錯(cuò)誤事實(shí)內(nèi)在相關(guān)性。本文提出了一種知識(shí)庫(kù)具有關(guān)系敏感嵌入式方法NSIL,以獲取知識(shí)庫(kù)各關(guān)系之間的相關(guān)性,從而檢查出知識(shí)庫(kù)中的錯(cuò)誤,,以此提高知識(shí)庫(kù)的準(zhǔn)確性與質(zhì)量。該方法分為相關(guān)性處理和錯(cuò)誤檢測(cè)兩階段,。在相關(guān)性處理階段,,使用NSIL的相關(guān)函數(shù)以分值形式獲取各關(guān)系之間的相關(guān)度;在錯(cuò)誤檢測(cè)階段,,基于相關(guān)度分值進(jìn)行錯(cuò)誤檢測(cè),,對(duì)于缺失主體或客體的三元組進(jìn)行缺失成分預(yù)測(cè)。最后在知識(shí)庫(kù)之一Freebase生成的基準(zhǔn)數(shù)據(jù)集“FB15K”上進(jìn)行了廣泛驗(yàn)證,,證明了該方法在知識(shí)庫(kù)錯(cuò)誤知識(shí)檢測(cè)方面有著很高的性能,。
中圖分類(lèi)號(hào): TP183
文獻(xiàn)標(biāo)識(shí)碼: A
DOI: 10.19358/j.issn.2096-5133.2020.10.005
引用格式: 繆琦,楊昕悅. 具有關(guān)系敏感嵌入的知識(shí)庫(kù)錯(cuò)誤檢測(cè)[J].信息技術(shù)與網(wǎng)絡(luò)安全,,2020,,39(10):23-27,37.
Knowledge base error detection with relation sensitive embedding
Miao Qi,,Yang Xinyue
School of Electronic and Information Engineering,,Liaoning Technical University,Huludao 125105,China
Abstract: Accuracy and quality are very important for the knowledge base. Although there have been many researches on the incompleteness of knowledge base, few workers consider the detection of errors in the knowledge base. According to the traditional methods, it is usually unable to effectively capture the internal correlation of errors in the knowledge base, so as to check the errors. In this paper, a relational sensitive embedded method NSIL for knowledge base is proposed to obtain the correlation among the relationships between them, so as to check out the errors in the knowledge base, so as to improve the accuracy and quality of the knowledge base. This method is divided into two stages: correlation processing and error detection. In the correlation processing stage, correlation function of NSIL is used to obtain the correlation degree of each relationship in the form of score; in the error detection stage, error detection is based on the score of correlation degree, and missing component prediction is carried out for the triplet of missing subject or object. At last, the method is verified on the benchmark data set "FB15K" which is generated by Freebase, one of the largest knowledge bases. It is proved that the method has high performance in knowledge base error detection.
Key words : knowledge base,;embedding model,;error detection

0 引言

    如今,知識(shí)庫(kù)已經(jīng)成為各種研究和應(yīng)用越來(lái)越重要的和常用的數(shù)據(jù)源,,如語(yǔ)義搜索,、實(shí)體鏈接、問(wèn)答系統(tǒng)和自然語(yǔ)言處理等,。為了使龐大數(shù)據(jù)庫(kù)更易于操作,,研究者提出了一種新的研究方向——知識(shí)庫(kù)嵌入。關(guān)鍵思想是嵌入KB(Knowledge Base)組件,,包括將實(shí)體和關(guān)系轉(zhuǎn)化為連續(xù)的向量空間,,從而簡(jiǎn)化操作,同時(shí)保留KB原有的結(jié)構(gòu),。實(shí)體和關(guān)系嵌入能進(jìn)一步應(yīng)用于各種任務(wù)中,,如KB補(bǔ)全、關(guān)系提取,、實(shí)體分類(lèi)和實(shí)體解析,。雖然龐大的知識(shí)庫(kù)中有數(shù)以億計(jì)的事實(shí),但是在信息爆炸的時(shí)代遠(yuǎn)遠(yuǎn)不夠,。大部分的研究工作聚焦知識(shí)庫(kù)對(duì)缺失邊的擴(kuò)充,,很少有人考慮到其中過(guò)時(shí)的、不正確的信息[1-3],。許多擴(kuò)充知識(shí)庫(kù)研究將事實(shí)投射到k維向量空間,通過(guò)聚類(lèi)來(lái)找到關(guān)系的相關(guān)性,,很難實(shí)現(xiàn)高效有效處理,。




本文詳細(xì)內(nèi)容請(qǐng)下載:http://forexkbc.com/resource/share/2000003133




作者信息:

繆  琦,楊昕悅

(遼寧工程技術(shù)大學(xué) 電子與信息工程學(xué)院,,遼寧 葫蘆島125105)

此內(nèi)容為AET網(wǎng)站原創(chuàng),,未經(jīng)授權(quán)禁止轉(zhuǎn)載。