第九章 結論與未來討論
9.2 未來研究方向
本論文運用詞彙關係分類模型自動分類查詢詞彙關係，並以關聯句比對規則 挑選模型自動計算句型分數，這些分類模型都需要事先蒐集大量資料並且由人工 標記類別。未來若能用採用像是部分監督式學習(partially supervised learning)的方 法，例如 self-training，先以少量且有標記的資料訓練分類模型，之後輸入未標 記的資料，逐次將分類模型自動標記的資料加入為訓練資料中，再重新訓練，如 此一來便可以降低人工標記類別的成本，並有持續學習的效果。此外我們認為可 以加入更多自然語言處理技術，像是語法樹(syntax tree)，藉以找出句子中最重要 的部分當作特徵，如此一來在計算語意關聯度上應該會有所提升；在挑選關聯句 組時，句型上的相似也可以是一種考量，未來可以考慮此種特徵來計算句子和句 子之間的相似性。本研究只考慮有出現查詢詞彙的句子，才會被當作是候選關聯 句集，未來可以考慮以多個連續句子或是段落為分析的基本單位，藉由更多的前 後文句或是字詞內容使得在計算語意關聯度時能更加準確。
66
參考文獻
[1] N. Schlaefer, J. Chu-Carroll, and E. Nyberg, ―Statistical Source Expansion for Question Answering,‖ in Proceedings of the 20th ACM conference on Information and Knowledge Management (CIKM), 2011.
[2] H. T. Dang, D. Kelly and J. Lin, ―Overview of the TREC 2007 Question Answering Track,‖ in Proceedings of the Sixteenth Text Retrieval Conference (TREC), 2007.
[3] X. Cao, G. Cong, and B. Cui, ―The Use of Categorization Information in Language Models for Question Retrieval,‖ in Proceedings of the 18th ACM conference on Information and Knowledge Management (CIKM), 2009.
[4] L. Cai, G. Zhou and K. Liu, "Large-Scale Question Classification in cQA by Leveraging Wikipedia Semantic Knowledge", in Proceedings of the 20th ACM conference on Information and Knowledge Management (CIKM), 2011
[5] Song, Y., Qiu, B., and Farooq, U. ―Hierarchical tag visualization and application for tag recommendations.‖ in Proceedings of the 20th ACM conference on Information and Knowledge Management (CIKM), 2011
[6] D. Bollegala, Y. Matsuo, and M. Ishizuka, "Measuring the SimilarityBetween Implicit Semantic Relations Using Web Search Engines", in Proceedings of the Second ACM International Conference on Web Search and Data Mining(WSDM), 2009.
[7] A. Kalyanpur, S. Patwardhan, and B. Boguraev, ―Fact-Based Question Decomposition for Candidate Answer Re-Ranking‖ in Proceedings of the 20th ACM conference on Information and Knowledge Management (CIKM), 2011 [8] X. Xue, J. Jeon, and W. B. Croft, ―Retrieval models for question and answer
archives,‖ in Proceedings of the 31rd Annual International ACM conference on Special Interest Group on Information Retrieval (SIGIR), 2008
[9] G. Luo, C. Tang, and Y. Tian, ―Answering relationship queries on the web‖ in Proceedings of the 16th international conference on World Wide Web(WWW), 2007
[10] S.E. Robertsom, S. Walker, and M. Hancock-Beaulieu, ―Okapi at TREC-7:
Automatic Ad Hoc, Filtering, VLC and Interactive‖, In proceedings of the 7th Text Retrieval Conference(TREC-7), NIST Special Publication.
[11] D. Jiang, K. W. Leung, W. Ng, ―Context-Aware Search Personalization with Concept Preference,‖ in Proceedings of the 20th ACM conference on Information and Knowledge Management (CIKM), 2011
[12] S. Szumlanski and F. Gomez, ―Automatically Acquiring a Semantic Network of
67
Related Concepts‖ in Proceedings of the 19th ACM conference on Information and Knowledge Management (CIKM), 2010
[13] C. Fellbaum, editor. ―WordNet: An electronic Lexical Database‖ MIT Press, 1998
[14] C. Fautsch, J. Savoy, ―Adapting the tf-idf Vector Space Model to Domain Specific Information Retrieval‖ in proceedings of 25th ACM Symposium on Applied Computing(SAC), 2010
[15] D. Vandic, J. V. Dam and F. Hogenboom, ―A Semantic Clustering-Based Approach for Searching and Browsing Tag Spaces‖ in proceedings of 26th ACM Symposium on Applied Computing(SAC), 2011
[16] M. S. Pera, R. Qumsiyeh, Y. K. Ng, ―A Query-Based Multi-document Sentiment Summarizer‖ in Proceedings of the 20th ACM conference on Information and Knowledge Management (CIKM), 2011.
[17] 謝聿承， 「兩個專有詞彙概念關聯句自動擷取技術之研究」 ，國立臺灣師 範大學，碩士論文，民國 100 年。
[18] H. Cui, M. Kan and T. Chua, ―Generic Soft Pattern Models for Definitional Question Answering‖ ACM Transactions on Information Systems, Vol. 25, No. 2, Article 8, April 2007.
[19] R.-E. Fan, P.-H. Chen, and C.-J. Lin. ―Working set selection using the second order information for training SVM,‖ Journal of Machine Learning Research 6, 1889-1918, 2005
68
附錄 A
包含關係訓練查詢字組
1 sort algorithm & bogosort 30 graph & spqr tree
2 sort algorithm & cocktail sort 31 algorithm & bareiss algorithm 3 sort algorithm & comb sort 32 algorithm & criss cross algorithm 4 sort algorithm & cycle sort 33 data structure & self organizing list 5 sort algorithm & gnome sort 34 data structure & doubly linked list 6 sort algorithm & introsort 35 data structure & unrolled linked list 7 sort algorithm & odd even sort 36 data structure & xor linked list 8 sort algorithm & patience sorting 37 data structure & aa tree
9 sort algorithm & smoothsort 38 data structure & scapegoat tree 10 sort algorithm & stooge sort 39 data structure & splay tree 11 sort algorithm & timsort 40 data structure & treap
12 list & self organizing list 41 data structure & weight balanced tree 13 list & doubly linked list 42 data structure & binary heap
14 list & unrolled linked list 43 data structure & bionomial heap 15 list & xor linked list 44 data structure & soft heap 16 tree & aa tree 45 data structure & skew heap 17 tree & scapegoat tree 46 data structure & incidence list 18 tree & splay tree 47 data structure & incidence matrix 19 tree & treap 48 data structure & skip graph 20 tree & weight balanced tree 49 data structure & scene graph 21 heap & binary heap 50 data structure & combinatorial map 22 heap & bionomial heap 51 data structure & spqr tree
23 heap & soft heap 52 clustering & hierarchical clustering 24 heap & skew heap 53 clustering & fuzzy clustering 25 graph & incidence list 54 clustering & k means
26 graph & incidence matrix 55 clustering & k medoids 27 graph & skip graph 56 clustering & k medians
28 graph & scene graph 57 hierarchical clustering & single linkage clustering 29 graph & combinatorial map 58 hierarchical clustering & complete linkage clustering
69
非包含關係訓練查詢字組
1 bogosort_cocktail sort 44 introsort_timsort
2 bogosort_comb sort 45 odd even sort_patience sorting 3 bogosort_cycle sort 46 odd even sort_smoothsort 4 bogosort_gnome sort 47 odd even sort_stooge sort 5 bogosort_introsort 48 odd even sort_timsort 6 bogosort_odd even sort 49 patience sorting_smoothsort 7 bogosort_patience sorting 50 patience sorting_stooge sort 8 bogosort_smoothsort 51 patience sorting_timsort 9 bogosort_timsort 52 smoothsort_stooge sort 10 cocktail sort_comb sort 53 smoothsort_timsort
11 cocktail sort_cycle sort 54 self organizing list_doubly linked list 12 cocktail sort_gnome sort 55 self organizing list_unrolled linked list 13 cocktail sort_introsort 56 self organizing list_xor linked list 14 cocktail sort_odd even sort 57 doubly linked list_unrolled linked list 15 cocktail sort_patience sorting 58 doubly linked list_xor linked list 16 cocktail sort_smoothsort 59 unrolled linked list_xor linked list 17 cocktail sort_stooge sort 60 aa tree_scapegoat tree
18 cocktail sort_timsort 61 aa tree_splay tree 19 comb sort_cycle sort 62 aa tree_treap
20 comb sort_gnome sort 63 aa tree_weight balanced tree 21 comb sort_introsort 64 scapegoat tree_splay tree 22 comb sort_odd even sort 65 scapegoat tree_treap
23 comb sort_patience sorting 66 scapegoat tree_weight balanced tree 24 comb sort_smoothsort 67 splay tree_treap
25 comb sort_stooge sort 68 splay tree_weight balanced tree 26 comb sort_timsort 69 treap_weight balanced tree 27 cycle sort_gnome sort 70 binary heap_bionomial heap 28 cycle sort_introsort 71 binary heap_soft heap 29 cycle sort_odd even sort 72 binary heap_skew heap 30 cycle sort_patience sorting 73 bionomial heap_soft heap 31 cycle sort_smoothsort 74 bionomial heap_skew heap 32 cycle sort_stooge sort 75 soft heap_skew heap
33 cycle sort_timsort 76 hierarchical clustering_fuzzy clustering 34 gnome sort_introsort 77 hierarchical clustering_k means
35 gnome sort_odd even sort 78 hierarchical clustering_k medoids 36 gnome sort_patience sorting 79 hierarchical clustering_k medians
70
37 gnome sort_smoothsort 80 fuzzy clustering_k means 38 gnome sort_stooge sort 81 fuzzy clustering_k medoids 39 gnome sort_timsort 82 fuzzy clustering_k medians 40 introsort_odd even sort 83 fuzzy clustering_k means 41 introsort_patience sorting 84 k means_k medoids 42 introsort_smoothsort 85 k means_k medians 43 introsort_stooge sort 86 k medoids_k medians
87 single linkage clustering_complete linkage clustering