知識(shí)圖譜的自然語(yǔ)言查詢和關(guān)鍵詞查詢

章 緒論·································．1
1．1 研究背景及意義··················．1
1．2 研究現(xiàn)狀···························．3
1．2．1 知識(shí)圖譜自然語(yǔ)言查詢的
研究現(xiàn)狀························3
1．2．2 知識(shí)圖譜關(guān)鍵詞查詢的
研究現(xiàn)狀························4
1．3 存在的關(guān)鍵問題··················．5
1．4 研究?jī)?nèi)容及創(chuàng)新點(diǎn)···············．7
1．5 本書組織結(jié)構(gòu)·····················10
第2 章 自然語(yǔ)言查詢和關(guān)鍵詞查詢的
相關(guān)研究···························12
2．1 知識(shí)圖譜的自然語(yǔ)言查詢······12
2．1．1 語(yǔ)義關(guān)系識(shí)別················．12
2．1．2 自然語(yǔ)言聚集查詢···········．13
2．1．3 查詢映射·····················．14
2．1．4 多樣化的自然語(yǔ)言查詢······．15
2．2 知識(shí)圖譜的關(guān)鍵詞查詢·········16
2．2．1 模式圖························．16
2．2．2 多樣化的關(guān)鍵詞查詢········．17
2．3 兩種查詢共用的基礎(chǔ)技術(shù)······19
2．3．1 實(shí)體識(shí)別和實(shí)體鏈接········．19
2．3．2 解釋詞典·····················．19
2．4 眾包—輔助語(yǔ)義關(guān)系識(shí)別規(guī)則
挖掘·································20
2．5 知識(shí)圖譜的其他非結(jié)構(gòu)化
查詢方式···························21
2．5．1 交互式查詢···················．21
2．5．2 實(shí)例查詢和樣例查詢········．22
第3 章 基于眾包的自然語(yǔ)言查詢中
語(yǔ)義關(guān)系識(shí)別規(guī)則挖掘·········23
3．1 問題描述及創(chuàng)新點(diǎn)···············23
3．2 眾包模型···························24
3．2．1 迭代模型和并行模型········．25
3．2．2 迭代式并行模型和
并行式迭代模型·············．25
3．2．3 帶反饋的并行式迭代模型···．26
3．3 生成語(yǔ)義關(guān)系數(shù)據(jù)集和
依賴結(jié)構(gòu)數(shù)據(jù)集··················27
3．3．1 眾包模型標(biāo)記語(yǔ)義關(guān)系·····．27
3．3．2 Stanford Parser 生成依賴
結(jié)構(gòu)··························．27
3．4 挖掘語(yǔ)義關(guān)聯(lián)規(guī)則···············28
3．4．1 挖掘語(yǔ)義關(guān)聯(lián)規(guī)則的算法···．28
3．4．2 算法MSAR 的復(fù)雜度·······．30
3．5 實(shí)驗(yàn)結(jié)果及分析—眾包模型··31
3．5．1 實(shí)驗(yàn)數(shù)據(jù)及評(píng)估標(biāo)準(zhǔn)········．31
3．5．2 迭代模型和并行模型········．32
3．5．3 迭代式并行模型和并行式
迭代模型·····················．33
3．5．4 帶反饋的并行式迭代模型···．35
3．6 實(shí)驗(yàn)結(jié)果及分析—語(yǔ)義關(guān)聯(lián)
規(guī)則·································36
3．7 語(yǔ)義關(guān)系識(shí)別·····················38
3．7．1 語(yǔ)義關(guān)系識(shí)別的算法········．38
3．7．2 算法SRR 的復(fù)雜度··········．39
3．7．3 實(shí)驗(yàn)結(jié)果及分析—語(yǔ)義關(guān)系
識(shí)別··························．39
3．8 本章小結(jié)···························40
第4 章 知識(shí)圖譜的自然語(yǔ)言聚集
查詢·································42
4．1 問題描述及創(chuàng)新點(diǎn)···············42
4．2 查詢流程···························45
4．3 查詢理解···························45
4．3．1 意圖解釋·····················．45
4．3．2 依賴結(jié)構(gòu)分類················．46
4．3．3 從依賴結(jié)構(gòu)中識(shí)別意圖解釋·．47
4．3．4 查詢理解的優(yōu)化·············．49
4．3．5 算法AIII 的復(fù)雜度··········．49
4．4 構(gòu)建基本圖模式··················50
4．4．1 擴(kuò)展的解釋詞典ED ·········．50
4．4．2 短語(yǔ)映射·····················．51
4．4．3 謂詞-類型鄰近集PT ·········．51
4．4．4 謂詞-謂詞鄰近集PP ·········．53
4．4．5 語(yǔ)義關(guān)系映射················．53
4．4．6 算法SRM 的復(fù)雜度·········．55
4．4．7 構(gòu)建基本圖模式BGP········．56
4．4．8 算法BBGP 的復(fù)雜度········．57
4．5 將基本圖模式翻譯為
SPARQL 語(yǔ)句·····················58
4．5．1 數(shù)值型謂詞···················．58
4．5．2 翻譯基本圖模式·············．59
4．5．3 翻譯聚集·····················．59
4．5．4 算法TA 的復(fù)雜度···········．61
4．6 實(shí)驗(yàn)結(jié)果及分析··················61
4．6．1 實(shí)驗(yàn)數(shù)據(jù)集···················．61
4．6．2 各階段的優(yōu)化能力···········．61
4．6．3 算法的有效性················．63
4．6．4 與現(xiàn)有算法對(duì)比·············．65
4．6．5 回答錯(cuò)誤的原因·············．66
4．7 相關(guān)問題及解決方案············67
4．8 本章小結(jié)···························69
第5 章 知識(shí)圖譜的自然語(yǔ)言查詢—
SPARQL 和關(guān)鍵詞··············70
5．1 問題描述及創(chuàng)新點(diǎn)···············70
5．2 查詢流程···························71
5．3 SPARQL 語(yǔ)句的生成過(guò)程······72
5．4 查詢分解···························73
5．4．1 查詢理解階段················．73
5．4．2 查詢映射階段················．74
5．4．3 執(zhí)行SPARQL 階段··········．74
5．4．4 查詢分解算法················．75
5．4．5 算法DQ 的復(fù)雜度···········．76
5．5 構(gòu)建關(guān)鍵詞索引··················77
5．5．1 算法QUKI ···················．77
5．5．2 算法QUKI 的復(fù)雜度········．78
5．6 聚合SPARQL 結(jié)果子圖和
關(guān)鍵詞查詢························78
5．6．1 算法CSK ····················．78
5．6．2 算法CSK 的復(fù)雜度·········．80
5．7 實(shí)驗(yàn)結(jié)果及分析··················81
5．7．1 算法的有效性················．81
5．7．2 回答正確的原因·············．83
5．7．3 回答錯(cuò)誤的原因·············．84
5．7．4 以SPARQL 查詢?yōu)橹鲗?dǎo)的
優(yōu)勢(shì)··························．85
5．7．5 關(guān)鍵詞索引的效率···········．85
5．8 本章小結(jié)···························86
第6 章 知識(shí)圖譜的關(guān)鍵詞聚集查詢···88
6．1 問題描述及創(chuàng)新點(diǎn)···············88
6．2 查詢流程···························90
6．3 構(gòu)建類型-謂詞圖·················90
6．3．1 關(guān)系提取·····················．90
6．3．2 關(guān)系標(biāo)準(zhǔn)化··················．91
6．3．3 類型-謂詞圖··················．92
6．4 查詢理解···························92
6．5 基于類型-謂詞圖構(gòu)建
查詢圖······························94
6．5．1 查詢圖························．94
6．5．2 構(gòu)建查詢圖··················．94
6．5．3 算法BQG 的復(fù)雜度·········．99
6．6 將查詢圖翻譯為SPARQL
語(yǔ)句·································99
6．6．1 數(shù)值型謂詞···················．99
6．6．2 翻譯一般路徑················．99
6．6．3 翻譯聚集·····················100
6．6．4 算法TQGS 的復(fù)雜度········102
6．7 實(shí)驗(yàn)結(jié)果及分析···············．102
6．7．1 算法的有效性················102
6．7．2 輸入的可擴(kuò)展性·············104
6．7．3 數(shù)據(jù)集的可擴(kuò)展性···········106
6．7．4 組件的有效性················106
6．8 本章小結(jié)························．108
第7 章 總結(jié)與展望·····················．109
7．1 總結(jié)······························．109
7．2 展望······························．111
參考文獻(xiàn)····································．112