應用探勘技術於社會輿情以預測捷運週邊房地產市場之研究 - 政大學術集成
69
0
0
全文
(2) ! !. !. !. 立立. 政 治 大. •‧. •‧ 國. ㈻㊫學. n. er. io. sit. y. Nat al. !. Ch. engchi. ii!. i Un. v.
(3) !. 99. 1. 1. 103. 250. 立立. 6. 30. 1,1150. 政 治 大. •‧. •‧ 國. ㈻㊫學. n. er. io. sit. y. Nat al. Ch. engchi 69.2. 78. !. 8,165. iii!. i Un. v.
(4) !. Abstract Nowadays, E-News have become an important way for people to get daily information. These enormous amounts of news could reflect public opinions on a particular attention or sentiment trends in news topics. Therefore, how to use opinion mining and sentiment analysis technology to dig out valuable information from particular news becomes the latest issue. In this study, we collected 1,1150 house news and 8,165 house transaction records. 政 治 大. around the MRT stations within 250 meters over the last five years. We extracted the. 立立. emotion words from the news by manipulating opinion mining. Furthermore, we built. •‧ 國. ㈻㊫學. moving average lines and the slope of the moving average in order to explore the. •‧. relationship and entry point between public opinion and housing market.. sit. y. Nat. In conclusion, we indicated that there is a high correlation between the news sentiment. io. n. al. er. and housing market. We also uses SVM algorithm to construct a model to predict housing. i Un. v. hotspots. The results demonstrate that the SVM model reaches average accuracy at 69.2%. Ch. engchi. and the model accuracy increases up to 78% for predicting housing hotspots. Besides, we also provide investors with a basis of entry point into the housing market by utilizing the moving average cross overs and slopes analysis and a better way of predicting housing hotspots .. Key Words. Text Mining, Opinion Mining, Housing Market, Moving Average,. Support Vector Machine. !. iv!.
(5) !. .......................................................................................................................................... II ABSTRACT............................................................................................................................ IV .......................................................................................................................................... V ................................................................................................................................... VII ..................................................................................................................................... IX. 政 治 大. ........................................................................................................................... 1. 立立. ......................................................................................................... 1. 1.2. ..................................................................................................................... 2. •‧ 國. ㈻㊫學. 1.1. •‧. ................................................................................................................... 4. 2.2. ................................................................................................. 5. al. n. 2.2.1 2.2.2 2.2.3. er. io. sit. y. ..................................................................................................................... 4. Nat. 2.1. iv 5 n C h................................................................................... U engchi ....................................................................................... 6 ............................................................................... 7. 2.3. ................................................................................................................. 8. 2.4. ............................................................................................................................. 9 ..................................................................................................... 10. !. 3.1. ....................................................................................................... 10. 3.2. ....................................................................................................... 11 v!.
(6) !. 3.2.1. ..................................................................................................... 11. 3.2.1. ......................................................................................................... 12. 3.3. ............................................................................................... 15. 3.3.1. ............................................................................................. 15. 3.3.2. ............................................................................................. 18. 3.4. ................................................................................... 21. 3.4.1. ..................................................................................................... 23. 3.4.2. 政 治 大 ..................................................................................................... 26 立立. ㈻㊫學. •‧ 國. 3.4.3. ................................................................................................. 24. ................................................................................................................. 30. ............................................................................................................ 30. 4.2. y. sit. n. al. er. .................................................................................................... 30. io. 4.1.2. Nat. 4.1.1. ................................................................................................... 30. •‧. 4.1. i Un. v. ................................................................................................... 44. Ch. engchi. 4.2.1. ............................................................................................................ 44. 4.2.2. ........................................................................................ 44 ..................................................................................................... 54. 5.1. ............................................................................................................... 54. 5.2. ....................................................................................... 55 ................................................................................................................................. 57. !. vi!.
(7) !. 3- 1. ................................................................................................ 10. 3- 2. ................................................................................................ 11. 3- 2. ................................................................................................ 21. 4- 1. ................................................................................................31. 4- 2. ................................................................................................ 31. 4- 3. ................................................................................................ 32. 4- 4. ................................................................................................ 32. 4- 5. ................................................................................................ 32. sit. y. ............................................................................................ 34. er. ............................................................................................ 34. •‧ 國. 4- 12. .......................................................................................... 35. 4- 13. .......................................................................................... 35. 4- 14. .......................................................................................... 35. 4- 15. .......................................................................................... 35. 4- 17. ...................................................................................... 36. 4- 18. .................................................................................. 36. n. 4- 11. a l.......................................................................................... iv 34 n Ch U engchi. 4- 10. !. io. 4- 9. ............................................................................................ 34. Nat. 4- 8. ............................................................................................ 34. •‧. 4- 7. ㈻㊫學. 4- 6. 立立. 政 治 大. .......................................................................................... 34. vii!.
(8) !. 4- 19. .............................................................................. 38. 4- 20. .......................................................................... 38. 4- 21. .............................................................................. 39. 4- 22. .......................................................................... 39. 4- 23. .............................................................................. 40. 4- 24. .......................................................................... 41. 4- 25. .............................................................................. 42. 4- 26. .......................................................................... 42 政 治 大 .............................................................................. 43. 立立. 4- 27. .......................................................................... 43. •‧ 國. ㈻㊫學. 4- 28. PRECISION-RECALL-CURVE ......................................................................... 45. 4- 30. ROC CURVE ................................................................................................ 45. sit. al. er. ...................................................................................................... 46. io. v. 4- 32. PRECISION-RECALL-CURVE ......................................................................... 47. 4- 33. ROC CURVE ................................................................................................ 47. 4- 34. ...................................................................................................... 48. 4- 35. PRECISION-RECALL-CURVE ......................................................................... 49. 4- 36. ROC CURVE ................................................................................................ 50. 4- 37. ...................................................................................................... 50. 4- 38. PRECISION-RECALL-CURVE ......................................................................... 51. 4- 39. ROC CURVE ................................................................................................ 52. 4- 40. ...................................................................................................... 52. n. !. y. Nat. 4- 31. •‧. 4- 29. Ch. engchi. viii!. i Un.
(9) !. 3- 1. ............................................................................................................ 12. 3- 2. ........................................................................................................ 13. 3- 3. ............................................................................................................ 15. 3- 4. ................................................................................................ 24. 3- 5. .................................................................................................................... 27. 4- 1. ................................................................................33. 4- 2. ................................................................................................ 37. 4- 3. .................................................................................................... 46. •‧ 國. .................................................................................................... 51 .................................................................................................... 53. n. al. er. io. sit. y. Nat. 4- 6. .................................................................................................... 48. •‧. 4- 5. ㈻㊫學. 4- 4. 立立. 政 治 大. !. Ch. engchi. ix!. i Un. v.
(10) !. 1.1. 立立. 政 治 大. •‧. •‧ 國. ㈻㊫學. n. er. io. sit. y. Nat al. !. Ch. engchi. 1!. i Un. v. ,2012.
(11) !. 立立. ,2006 政 治 大. •‧. •‧ 國. ㈻㊫學. n. er. io. sit. y. Nat al. Ch. engchi. 1.2. !. 2!. i Un. v.
(12) !. 1. 2. 3.. 立立. 政 治 大. •‧. •‧ 國. ㈻㊫學. n. er. io. sit. y. Nat al. !. Ch. engchi. 3!. i Un. v.
(13) !. 2.1 Opinion Mining. Sentiment Analysis. Pang and Lee,2008. 政 治 大. 2014. 立立. •‧ 國. 2010. Pang & Lee 2005. Mullen & Collier. 2004. 2007. •‧ sit. y. Nat io. n. er. 2012. al. Ch. engchi. i Un. v. Wuthrich 1998. 2011 2014. Bollen !. 2004. ㈻㊫學. Ku & Chen 2007. Hu & Liu. Johan et al. 2010 4!. Twitter.
(14) !. 86.7%. 2012. 2014. 2012. 2.2. 立立. 政 治 大. •‧. •‧ 國. ㈻㊫學. n. er. io. sit. y. Nat al. v. Hickling Lewis Brod, Inc. 2002. Ch. engchi. i Un. 1983. 2009. 2012. !. 5!.
(15) !. ,2003 Tesarek,2003. Smith &. 2004. 立立. 2013 政 治 大. •‧. •‧ 國. ㈻㊫學. n. er. io. sit. y. Nat al. Ch. engchi U. v ni. 1994. 2012 1999. 2009. 2007 250 !. 6!.
(16) !. 200. Public Opinion. 立立. 政 治 大. •‧ 國. ㈻㊫學. Bounded Rationality. Duncan&Sheridan. Simon,1982. •‧. n. er. io. sit. y. Nat al. 2006. Ch. engchi. i Un. v. Chinloy 1996 Shiller 2007. 2007 2006. !. 7!. 2007.
(17) !. 2012. 2012. 立立. 2.3. 政 治 大 Vapnik 1995. •‧. •‧ 國. ㈻㊫學. Support Vector Machine, SVM. n. er. io. sit. y. Nat al. Ch. engchi. i Un. v. Quadratic. Programming ,2009. Wei Huang et al 2005 2008. 2009 !. 8!.
(18) !. 2003. 2012 20. 2012. 立立. •‧. •‧ 國. ㈻㊫學. 2.4. 政 治 大. n. er. io. sit. y. Nat al. !. Ch. engchi. 9!. i Un. v. 72.2.
(19) !. 政 治 大. 3.1. !. !. !. !. n. al. !. !. Ch. e n g c! h i. !. !. 3- 1. !. sit. io. !. !. !. y. Nat. !. NTUSD/. •‧. !. !. ! 250m. !. er. udn. 3- 1. ㈻㊫學. •‧ 國. 立立. 10!. i Un. v. !. !. ! !. !. !. 3- 2.
(20) !. SVM!Model!. !. ! !. !. !. ! !. ! !. !. !. !. SVM. !. !. !. 治 政 ! 大. 立立. •‧ 國. •‧. n. al. er. io. sit. y. Nat. 3.2. ㈻㊫學. 3- 2. Ch. engchi. i Un. v. 24 18. 6. 7. 62. 1.. 99 !. 1. 1. 101 11!. 7. 30. 7.
(21) !. 101 101. 8. 1. 103. 6. 30. 2.. Upaper. 立立. •‧ Context. MRT. y. Title. n. al. er. io. sit. Data Time. Ch. engchi. 1.. !. 3- 1. 3- 1. Nat. Id. •‧ 國. Html. Crawler. ㈻㊫學. XPath. Python 政 治 大. 12!. i Un. v. Udn.
(22) !. 2. python. jieba. jieba. Viterbi Hidden Markov Mode, HMM. jieba Vd i 政 治 大. ICTCLAS. 立立. 3- 2. ㈻㊫學. 10. 5. •‧. •‧ 國. 3- 2. 250. :n,. :n, :n,. :n, :n,. :f, :a,. :n,10:m,. Ch. :q,. :n,. v ni :n,. e n g:f,250:m, c h i U:q,. :n,. :n,10:m,. :q,. 3.. Stop Words !. 10. sit al. n. :d,. :n,. er. io :v,. z. y. Nat :n,. s. 13!. :n,. :n,. :ad,. :n, :a. :a,. :q,. :v,. :n,. :n, :a,.
(23) !. 4. National Taiwan University Sentiment Dictionary, NTUSD General Inquirer GI. Chinese Network Sentiment Dictionary CNSD. 11086. 立立. HowNet. 2810. 8276. 政 治 大 9193 1254. •‧ 國. ㈻㊫學. 836. 3730. •‧. 3116. Nat. 38. n. al. er. io. sit. y. 219. Ch. engchi. i Un. v. National Taiwan University Sentiment Dictionary, NTUSD. HowNet NTUSD. NTUSD. !. 14!.
(24) !. 3.3. 1. NTUSD. HowNet. , 2012 2011. 政 治 大 3- 3 ADV+V. •‧. V+V. V+N ADV+N. n. al. Ch. i Un. i e n g c h2011. 15!. er. io. sit. y. N+V. Nat. !. 3- 3. ㈻㊫學. •‧ 國. 立立. 80%. v.
(25) !. 2.. 5,334. 4,460. , 2012 PSO = NSO =. 立立. !!!!. w in P. (1). !!P !. 政 治 大 w in N !. !!!!. !"#. SO = log !"#. (3). •‧. •‧ 國. (2). !!!N. ㈻㊫學. Positive Sentiment Orientation. Nat io. sit. y. Dw in P. DP. n. al. PSO. Ch. Dw in N. NSO. engchi. er. PSO. !. v. Negative Sentiment Orientation. i Un. DN NSO. PSO. !. NSO. log. 16!. SO. SO. 1 SO. 1. SO. 1.
(26) !. SO. 1. -1. 0 !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 1!, !"!!!" = 0!!!"#!!!! !!!"!!"#$%$&'!!"#$%$&!!!!!!!!! SO = 0!, !"!!!" = 0!!!"#!!!! !"#!!"!!"#$#!$!!"#$%$& !! −1!, !"!!!" = 0!!!"#!!!! !!!"!!"#$%&'"!!"#$%$!!!!! !!!!!!!. 立立. 政 治 大. •‧. •‧ 國. ㈻㊫學. io. Document_Score Dj. n. al. ! ! Opinion_Score. !i. =. y. = !". iv n ! C +!3 ∗h ! Opinion_ScoreU !i ∈ !"!#$ engchi. Opinion_Score Oi. (6). Document_Score Dj. 1 -1 0 !. (5). sit. Nat. Opinion_Score Oi. er. 3.. (4). 17!.
(27) !. 4.. 政 治 大. 1.. 立立. •‧. •‧ 國. ㈻㊫學. n. al. er. io 250m. sit. y. Nat. 2.. Ch. engchi. i Un. v. 300 250. 250. 500. 550 “. 412. 15. 1-30. “ 250. 50. !. 18!.
(28) !. ASP.NET C#. Google Maps Javascript API. API. HTTP. Φ. Geocoder. λ. Great Circle Distance. Sinnott,1984 250~300. D = 2!×!!"#$%&. !!!!!. sin!. !. (7). !. Φ2,λ2. d. •‧. •‧ 國. ㈻㊫學. n. al. er. io. sit. y. Nat. 3.. !. 政 治 大 Φ1,λ1. 立立. R. !!!!". + cos ϕ1 cos ϕ2 sin!. Ch. engchi. i Un. v. 1.. Simple Moving Average, SMA. N Simple Average !. 19!.
(29) !. 1997 SMA_t =. P_t + P_[t-1] + P_[t-2] +…+ P_[t-N+1]. N. /N. t. (8). P. P_[t]. 2.. 立立. 政 治 大 Pearson Correlation Coefficient. •‧ 國. ㈻㊫學. !! ! − !. !. !! − !. •‧. !! ! − ! !! ! − !. !=. !. n. al. er. io. sit. y. Nat. 1.. (9). Ch. engchi. i Un. v. Double Moving Average, DMA. 2010 N. DMA_t =. !. SMA_t +SMA_[t-1] +… +SMA_[t-n+1]. 20!. /N. (10).
(30) !. 2.. 2004. 3.4. 政 治 大 Support Vector 立立 Machine, SVM. •‧. •‧ 國. ㈻㊫學. Hyper Plane. 3- 3. n. er. io. sit. y. Nat al. Ch. i Un. v. Separating Hyper Plane!. engchi. 3- 3. !. Separating. 21!. Marge!.
(31) !. f(x)= wx+b. yi= -1. f(x)> 0. f(x)< 0. yi= 1. H1: wx+b+1 = 0 H2: wx+b-1 = 0. (Margin) SVM. :. 1 Min! !wT!w 2. (11). 政 治 大. subject to yi (w ⋅ x + b)− 1 ≥ 0 , i = 1...n. ㈻㊫學. •‧ 國. 立立. ξ (Slack Variables). •‧. C. n. Ch. !. ξ. er. io. al. sit. y. Nat. 1 in! !wTw + C 2. v ni. !!!. (12). 1 . . .n e n gξ ≥c0h, ii = U. subject to yi (w ⋅x + b)+ ξ ≥ 1. Kernel Function Polynomial. RBF. Radius Basis Function. SVM. !. 22!. S. Linear Sigmoid.
(32) !. 1. (1). 500. (2). 500. 立立. sit. 500. n. al. 500. Ch. (8). 500. (9). 500. (10). er. io. (7). y. 500. Nat. (6). 500. •‧. (5). •‧ 國. (4). 500. ㈻㊫學. (3). 政 治 大. engchi. 500. (11) (12) (13) (14). !. 23!. i Un. v.
(33) !. 2.. (1) (2). 立立. Google Map. •‧. •‧ 國. ㈻㊫學. Google+. 政 治 大 Google Map. FonFood. n. er. io. sit. y. Nat al. 102. Ch 11. engchi 103. i Un. v. 10 3- 4. 3- 4. 500. !. Google Map. 24!.
(34) !. 500. Google Map. 500. Google Map FonFood. 500. Google Map. 500. Google Map. 500. Google Map. 500. Google Map. 500. Google Map. 500. FonFood. •‧. n. al. er. io. sit. y. Nat. !. Google Map. ㈻㊫學. •‧ 國. 立立500. 政 治 大. Ch. engchi. 25!. i Un. v.
(35) !. Z-score. 1.. Kernel Function. 立立. Radial Basis Function, RBF. •‧ 國. C. gamma. •‧. Hsu et al, 2003. n. al. er. io. sit. y. Nat. 2.. SVM. ㈻㊫學. RBF. 政 治 大. Ch. engchi. RBF. i Un. v. C gama. C gama. gama. Grid Search. !. 26!.
(36) !. 3. Over-Fitting Cross-Validation. Training. Data. Testing Data K-. K-Fold. Cross-Validation. -1. 政 治 大. K. 立立. •‧ 國. ㈻㊫學. 4.. Accuracy. •‧. Confusion Matrix. 3- 5. Precision. Recall. Nat. ROC Curve. n. al. er. io. sit. y. F1-Score. Ch. i Un. engchi. 3- 5. v. Relevant Retrieved. Relevant. Retrieved Not Retrieved. !. True Positives False Negatives. 27!. TP FN. NonRelevant False Positives. FP. True Negatives. TN.
(37) !. (1). Accuracy Wikipedia. TP + TN !! TP + FP + FN + TN !!. Accuracy =. (2). (12). Precision. 政 治 大. 立立. TP !! TP + FP !!. •‧. •‧ 國. ㈻㊫學. Precision =. sit. n. al. er. io. Recall. y. Nat. (3). Recall =. Ch. engchi. TP !! TP + FN !!. (4) F1-Score. !. (13). 28!. i Un. v. (14).
(38) !. Harmonic Mean. F1-Score F1 − Score =. !!!2 ∗ Precision ∗ Rcall!!! Precision + Rcall. (15). (5) ROC Curve Receiver Operating Characteristic Curve : FPR Y ROC. 政 治 大. FPR =. !!!FP!!! FP + TN. y sit. io. al. n. !. •‧. !!!TP!!! TP + FN. Nat. TPR =. ㈻㊫學. •‧ 國. 立立. 0, 0. er. 1,1. TPR. Ch. engchi. 29!. i Un. v. (16) (17).
(39) !. 4.1. 4.1.1. 立立. •‧ 國. y 1733. Ch. v ni. 2296. engchi U. 4.1.2. 1.. !. 1,1150. sit al. n. 2142. 9799. 30!. er. io. 982. 232. •‧. Nat. 250. 333. 150. ㈻㊫學. 413. 政 治 大 223. 1012 8165.
(40) n. al. er. sit. io. !. y. •‧. •‧ 國 立立. 4- 1. Ch. ㈻㊫學. Nat. 160! 140! 120! 100! 80! 60! 40! 20! 0! 099/1! 099/3! 099/5! 099/7! 099/9! 099/11! 100/1! 100/3! 100/5! 100/7! 100/9! 100/11! 101/1! 101/3! 101/5! 101/7! 101/9! 101/11! 102/1! 102/3! 102/5! 102/7! 102/9! 102/11! 103/1! 103/3! 103/5!. 0!. 099/1! 099/3! 099/5! 099/7! 099/9! 099/11! 100/1! 100/3! 100/5! 100/7! 100/9! 100/11! 101/1! 101/3! 101/5! 101/7! 101/9! 101/11! 102/1! 102/3! 102/5! 102/7! 102/9! 102/11! 103/1! 103/3! 103/5!. !. 4- 1 4- 5. 200!. 150!. 100!. 50!. 政 治 大. engchi. 4- 2. 31!. i Un. v.
(41) n. al. er. y. 立立. 4- 4. Ch. sit. •‧. io. !. ㈻㊫學. Nat. 160! 140! 120! 100! 80! 60! 40! 20! 0!. •‧ 國. 140! 120! 100! 80! 60! 40! 20! 0! 099/1! 099/3! 099/5! 099/7! 099/9! 099/11! 100/1! 100/3! 100/5! 100/7! 100/9! 100/11! 101/1! 101/3! 101/5! 101/7! 101/9! 101/11! 102/1! 102/3! 102/5! 102/7! 102/9! 102/11! 103/1! 103/3! 103/5!. 099/1! 099/3! 099/5! 099/7! 099/9! 099/11! 100/1! 100/3! 100/5! 100/7! 100/9! 100/11! 101/1! 101/3! 101/5! 101/7! 101/9! 101/11! 102/1! 102/3! 102/5! 102/7! 102/9! 102/11! 103/1! 103/3! 103/5!. 0!. 099/1! 099/3! 099/5! 099/7! 099/9! 099/11! 100/1! 100/3! 100/5! 100/7! 100/9! 100/11! 101/1! 101/3! 101/5! 101/7! 101/9! 101/11! 102/1! 102/3! 102/5! 102/7! 102/9! 102/11! 103/1! 103/3! 103/5!. !. 200!. 150!. 100! 50!. 4- 3. 政 治 大. engchi. 4- 5. 32!. i Un. v.
(42) !. 2.. Pearson Correlation Coefficient. 4- 1. 4- 1 3. 立立. /. 6. /. /. ㈻㊫學. 0.741. 0.690. 0.817. 0.679. 0.607. 0.754. 0.733. 0.779. 0.653. 0.862. 0.692. 0.618. 0.766. 0.740. n. Ch. 0.681. engchi. er. io. al. 0.818. sit. y. Nat. 0.624. •‧. •‧ 國. /. 政 治 大. i Un. v. 0.884. 0.730 0.760. 4- 6. !. 33!. 4- 14.
(43) 70!. 60!. al Ch engchi. 50!. 40!. 54! 52! 50! 48! 46! 44! 42! 40!. 4- 10. 34!. 4- 11. 101/9!. 102/12!. 102/9!. 102/6!. 102/3!. 101/12!. i Un. 101/6!. sit. y. 4- 9. 101/3!. n. 80! 40!. 100/12!. 40!. 立立. er. 50!. 100/9!. 60!. 100/6!. 70!. 100/3!. 4- 6. 099/12!. 80! 40! 099/6! 099/9! 099/12! 100/3! 100/6! 100/9! 100/12! 101/3! 101/6! 101/9! 101/12! 102/3! 102/6! 102/9! 102/12! 103/3! 103/6!. 40!. 099/6! 099/9! 099/12! 100/3! 100/6! 100/9! 100/12! 101/3! 101/6! 101/9! 101/12! 102/3! 102/6! 102/9! 102/12! 103/3! 103/6!. 50!. 099/9!. 4- 8. •‧. •‧ 國. 60!. ㈻㊫學. io. ! 099/3! 099/6! 099/9! 099/12! 100/3! 100/6! 100/9! 100/12! 101/3! 101/6! 101/9! 101/12! 102/3! 102/6! 102/9! 102/12! 103/3! 103/6!. 80!. 099/6!. 30! 099/3! 099/6! 099/9! 099/12! 100/3! 100/6! 100/9! 100/12! 101/3! 101/6! 101/9! 101/12! 102/3! 102/6! 102/9! 102/12! 103/3! 103/6!. 30!. Nat. 30!. 099/3! 099/6! 099/9! 099/12! 100/3! 100/6! 100/9! 100/12! 101/3! 101/6! 101/9! 101/12! 102/3! 102/6! 102/9! 102/12! 103/3! 103/6!. !. 70!. 55!. 50!. 45!. 4- 7. 60!. 55!. 政 治 大 50!. 45!. v.
(44) 60! 55!. 50!. 45!. 30! 40!. 4- 12. 70!. 50!. 40!. 立立. •‧. io. n. al Ch. 35! 40!. 4- 15. engchi. i Un. y. 60!. sit. 70!. 099/6! 099/9! 099/12! 100/3! 100/6! 100/9! 100/12! 101/3! 101/6! 101/9! 101/12! 102/3! 102/6! 102/9! 102/12! 103/3! 103/6!. 60!. 099/6! 099/9! 099/12! 100/3! 100/6! 100/9! 100/12! 101/3! 101/6! 101/9! 101/12! 102/3! 102/6! 102/9! 102/12! 103/3! 103/6!. 40!. 099/3! 099/6! 099/9! 099/12! 100/3! 100/6! 100/9! 100/12! 101/3! 101/6! 101/9! 101/12! 102/3! 102/6! 102/9! 102/12! 103/3! 103/6!. 80!. ㈻㊫學. 4- 14. Nat. !. •‧ 國. 80!. er. 30! 099/3! 099/6! 099/9! 099/12! 100/3! 100/6! 100/9! 100/12! 101/3! 101/6! 101/9! 101/12! 102/3! 102/6! 102/9! 102/12! 103/3! 103/6!. !. 50!. 4- 13. 55!. 50!. 政 治 大 45!. v.
(45) !. 3.. 4- 16. 4- 17. 政 治 大. 立立. •‧. •‧ 國. ㈻㊫學. 4- 16. n. er. io. sit. y. Nat al. Ch. engchi. 4- 17. !. 36!. i Un. v.
(46) !. SPSS. T. 4- 2. !0. ρ= 0. !1. ρ≠0. 4- 2. 政 治 大. .748**. .742**. .845**. .752**. .739**. .739**. •‧. .863**. ㈻㊫學. .686**. •‧ 國. 立立. .769**. Nat. y. .847**. n. al. er. io. sit. *p< .05 **p< .01 ***p< .001. Ch. engchi. i Un. v. (1) 0.686 !0. 0.01. !1. 4- 18. 101. 12 102. 11 !. 37!.
(47) n. !0. al. er. sit. y. 立立. 099/11! 100/1! 100/3! 100/5! 100/7! 100/9! 100/11! 101/1! 101/3! 101/5! 101/7! 101/9! 101/11! 102/1! 102/3! 102/5! 102/7! 102/9! 102/11! 103/1! 103/3! 103/5!. !0. 4- 18. ㈻㊫學. io. !. Nat. 80! 70! 60! 50! 40! 30! 20! 10! 0!. •‧. •‧ 國. 80! 70! 60! 50! 40! 30! 20! 10! 0!. 099/11! 100/1! 100/3! 100/5! 100/7! 100/9! 100/11! 101/1! 101/3! 101/5! 101/7! 101/9! 101/11! 102/1! 102/3! 102/5! 102/7! 102/9! 102/11! 103/1! 103/3! 103/5!. !. 0.863. 4- 19 101. Ch. engchi. 0.748. !1. 38!. 0.01. !1 8 102. i Un. 11. 政 治 大. v. 4- 19. (2). 0.01.
(48) y. !0. sit. •‧ 國. 4- 20. 4- 21. 立立. n. 4- 20. al. er. 099/11! 100/1! 100/3! 100/5! 100/7! 100/9! 100/11! 101/1! 101/3! 101/5! 101/7! 101/9! 101/11! 102/1! 102/3! 102/5! 102/7! 102/9! 102/11! 103/1! 103/3! 103/5!. •‧. io. !. Nat. 70! 60! 50! 40! 30! 20! 10! 0!. ㈻㊫學. 70! 60! 50! 40! 30! 20! 10! 0!. 099/11! 100/1! 100/3! 100/5! 100/7! 100/9! 100/11! 101/1! 101/3! 101/5! 101/7! 101/9! 101/11! 102/1! 102/3! 102/5! 102/7! 102/9! 102/11! 103/1! 103/3! 103/5!. !. 101. 101. Ch. 10 102. 0.742. !1 8 102. 政 治 大. engchi. 4- 21. 39!. i Un. 10. v. 11. 0.01.
(49) y. •‧. sit. !0 !1. 立立. n. al. er. 0.01 !0. 4- 22 101. 101. Ch 6 102. engchi. 4- 22. 40!. ㈻㊫學. io. 80! 70! 60! 50! 40! 30! 20! 10! 0!. Nat. !. •‧ 國. 4- 23. 099/11! 100/1! 100/3! 100/5! 100/7! 100/9! 100/11! 101/1! 101/3! 101/5! 101/7! 101/9! 101/11! 102/1! 102/3! 102/5! 102/7! 102/9! 102/11! 103/1! 103/3! 103/5!. !. (3) 0.845. !1 10 102. 0.752. 政 治 大. i Un. 0.01. 5. v. 11.
(50) n. !. y. 立立. !0. 4- 24. 101. al Ch. 101 11. 0.739. engchi. 41!. •‧. io. 4- 25. sit. (4) 0.739. !1. ㈻㊫學. Nat. !0. 099/11! 100/1! 100/3! 100/5! 100/7! 100/9! 100/11! 101/1! 101/3! 101/5! 101/7! 101/9! 101/11! 102/1! 102/3! 102/5! 102/7! 102/9! 102/11! 103/1! 103/3! 103/5!. 80! 70! 60! 50! 40! 30! 20! 10! 0!. er. •‧ 國. !. 4- 23. 政 治 大. 102. i Un. !1. 6. 1029. 0.01. 11. v 0.01.
(51) n. !. sit. io. 4- 26 4- 25. y. ㈻㊫學. Nat. !0. al. er. 80! 70! 60! 50! 40! 30! 20! 10! 0!. 099/11! 100/1! 100/3! 100/5! 100/7! 100/9! 100/11! 101/1! 101/3! 101/5! 101/7! 101/9! 101/11! 102/1! 102/3! 102/5! 102/7! 102/9! 102/11! 103/1! 103/3! 103/5!. 099/11! 100/1! 100/3! 100/5! 100/7! 100/9! 100/11! 101/1! 101/3! 101/5! 101/7! 101/9! 101/11! 102/1! 102/3! 102/5! 102/7! 102/9! 102/11! 103/1! 103/3! 103/5!. 80! 70! 60! 50! 40! 30! 20! 10! 0!. •‧. •‧ 國. !. 4- 24. 立立 政 治 大. Ch. (5). engchi. 101. 42!. i Un. v. 0.847 0.01. !1. 11. 102. 11.
(52) sit. y. 4- 27. n. al. er. 立立. 099/11! 100/1! 100/3! 100/5! 100/7! 100/9! 100/11! 101/1! 101/3! 101/5! 101/7! 101/9! 101/11! 102/1! 102/3! 102/5! 102/7! 102/9! 102/11! 103/1! 103/3! 103/5!. !0. 4- 26. ㈻㊫學. io. !. Nat. 80! 70! 60! 50! 40! 30! 20! 10! 0!. •‧. •‧ 國. 80! 70! 60! 50! 40! 30! 20! 10! 0!. 099/11! 100/1! 100/3! 100/5! 100/7! 100/9! 100/11! 101/1! 101/3! 101/5! 101/7! 101/9! 101/11! 102/1! 102/3! 102/5! 102/7! 102/9! 102/11! 103/1! 103/3! 103/5!. !. 0.769. 101. Ch. engchi. 4- 27. 43!. 0.01. !1 8 102. i Un. 10. 政 治 大. v.
(53) !. 4.2. 4.2.1 -. 政 治 大 14. -. 立立. 65. -. ㈻㊫學. •‧ 國. 103. 11. 15. •‧ y. Nat. 13. n. al. 8. er. io. sit. -. -. Ch. engchi. i Un. v. 4.2.2 1. 57. -. 8 Z-score. Grid Search C. 100 66%. !. Linear Accuracy Recall. 62.5%. 62%. Precision F1-Score. 44!. 63%.
(54) !. Precision-Recall-Curve. 4- 28. ROC Curve. 4- 28. 治 政Precision-Recall-Curve 大. 立立. 4- 29. •‧. •‧ 國. ㈻㊫學. n. er. io. sit. y. Nat al. Ch. engchi. 4- 29. i Un. v. ROC Curve. Confusion 4- 30. !. 45!. atrix.
(55) !. 50%. 67% 60%. F1-Score. 67%. 4- 3. 政 治 大. ㈻㊫學. •‧ 國. 立立. 57%. 4- 30. •‧. io. sit. y. Nat. 4- 3. n. al. er. 75%. F1-Score. i vF1-score n U. Precision. Recall. 0.50. 0.67. 0.57. 0.60. 0.67. Ch. engchi. 0.75. 2. 57. -. !. 46!. 8.
(56) !. Grid Search gamma. 0.001. Precision 63%. Accuracy 66%. Recall. Precision-Recall-Curve. 立立. 4- 31. 62% ROC Curve. •‧. •‧ 國. ㈻㊫學. Nat. sit. n. er. io. al. Precision-Recall-Curve. Ch. engchi. 4- 32. !. F1-Score. 政 治 大. 4- 31. i Un. ROC Curve. 47!. C. 62.5%. y. 10000.0. RBF. v. 4- 32.
(57) !. 4- 33 67% 60%. 50%. F1-Score. F1-Score. 57%. 75%. 67%. 4- 4. 立立. 政 治 大. •‧. •‧ 國. ㈻㊫學 y. n. al. er. io. sit. Nat. 4- 33. Ch 4- 4. !. engchi. i Un. v. Precision. Recall. F1-score. 0.50. 0.67. 0.57. 0.75. 0.60. 0.67. 48!.
(58) !. 3. 52. 13. -. Z-score. Grid Search Linear. Accuracy Recall Curve. 4- 34. 61.5%. Precision. 62%. F1-Score. ROC Curve. 立立. 63%. 4- 35. 政 治 大. •‧. •‧ 國. ㈻㊫學 y sit. n. er. io. Ch. engchi. i Un. v. Precision-Recall-Curve. 49!. 1.0 64%. Precision-Recall-. Nat al. 4- 34. !. C.
(59) !. 4- 35. io. n. al. 71%. Ch. engchi. 4- 36. !. 4- 5. er. F1-Score. 44%. y. F1-Score. sit. 45%. •‧. 67%. Confusion. 4- 36. Nat. 75%. 政 治 大. ㈻㊫學. 40%. •‧ 國. 立立. ROC Curve. 50!. i Un. v. atrix.
(60) !. 4- 5 Precision. Recall. F1-score. 0.40. 0.45. 0.44. 0.75. 0.67. 0.71. 4.. 政 治 大. 52. 立立. 4- 37. 69%. ROC Curve. io. 4- 38. n. al. 4- 37. !. 69%. y. F1-Score. er. Curve. Recall. Nat. 69%. 69.2%. •‧. Accuracy. Linear C. sit. Search. Grid. ㈻㊫學. •‧ 國. -. 13. Ch. engchi. i Un. v. Precision-Recall-Curve. 51!. 100. Precision Precision-Recall-.
(61) !. 4- 38. io. n. al. 78%. Ch. engchi. 4- 39. !. 4- 6. er. F1-Score. 50%. y. F1-Score. sit. 50%. •‧. 78%. Confusion. 4- 39. Nat. 78%. 政 治 大. ㈻㊫學. 50%. •‧ 國. 立立. ROC Curve. 52!. i Un. v. atrix.
(62) !. 4- 6 Precision. Recall. F1-Score. 0.5. 0.5. 0.5. 0.78. 0.78. 0.78. F1-Score Precision-Recall-Curve. 立立. 政 治 大. 69.2. F1-Score. •‧. •‧ 國. ㈻㊫學. n. al. er. io. sit. y. Nat. !. ROC Curve. Ch. engchi. 53!. i Un. v.
(63) !. 5.1. 立立. 政 治 大. •‧ 國. ㈻㊫學. 70%. t. •‧. n. er. io. sit. y. Nat al. !. Ch. engchi. 54!. i Un. v.
(64) !. 政 治 大. 立立. •‧. •‧ 國. ㈻㊫學. n. er. io. sit. y. Nat al. Ch. engchi. 5.2. 1. udn. !. 55!. i Un. v.
(65) !. 2.. 立立. •‧. •‧ 國. ㈻㊫學. 3.. 政 治 大. n. er. io. sit. y. Nat al. 4.. !. Ch. engchi U. 56!. v ni. !.
(66) !. , 2007, 18. ,. p1~26. , 2012,. -. ,2010,. ,. ,. ,2003,. ,. , 1999,. ,. , 2010,. 立立. •‧ 國 ,. ,. , 2003,. y. n. al. Ch. -. engchi. , 2012,. i Un. ,. , ─. -. ,. , 2006,. ,. , 2012,. !. ,. v. , 2009,. ,2005,. ,. sit. io. , 2013,. ,p.75-97. er. , 2003,. ,. ,. Nat. ,. ,. •‧. , 1993,. -. ㈻㊫學. , 2012,. , 政 治 大. ,. 57!. ,.
(67) !. , 2006,. ,. , 1994,. ,. pp.25-45 , 2006,. ,. , 2007,. ,. , 2009,. ,. , 2014,. 政 治 大. 立立SVM. , 2012,. •‧ 國. •‧ y. ,. sit. SVM. ,. Nat. ,2009,. ,. ㈻㊫學. , 2012,. ,. er. io. B. Wuthrich , V. Cho , S. Leung , D. Permunetilleke , K. Sankaran , J. Zhang , W. Lam.. al. 1998. Daily Stock Market Forecast from Textual Web Data. IEEE INTERNATIONAL. n. iv n C CONFERENCE ON SYSTEMS, MAN, CYBERNETICS h e nAND gchi U. Bo Pang & Lillian Lee. 2005. Seeing stars: Exploiting class relationships for sentiment categorization with respect to rating scales. Proceedings of the Association for Computational Linguistics. ACL , 115–124.. Bo Pang & Lillian Lee. 2008. Opinion mining and sentiment analysis. Foundations and Trends in Information Retrieval, 2. 1-2. , 1-135.. Bollen, Johan, Huina Mao, and Xiao-Jun Zeng. 2010. Twitter mood predicts the stock market. arXiv preprint arXiv:1010.3003. C. Cortes and V. Vapnik.1995. Support vector networks. Machine Learning, 20:125 Derek J. de Solla Price. 1963. Little science, big science. New York: Columbia University !. 58!.
(68) !. Press. Duncan, J. W., & Sheridan, P. D. 2007. Influentials, networks, and public opinion formation. Journal of Consumer Research, 34. 4. , 441-58.. Frame, Baum, Card. 1979. An information approach to examining developments in an energy technology: Coal gasification. Journal of the American Society for Information Science. Hickling Lewis Brod, Inc. 2002 , Commercial Property Benefits of Transit , Transportation Research Board,pp.1-48. Howard Rheingold.1993. The Virtual Community: Homesteading on the Electronic Frontier. London: MIT. 政 治 大. Hsu, C. W., Chang, C. C., Lin, C. J. 2003. A Practical Guideto Support Vector,. 立立. Classification. Available at http://www.csie.ntu.edu.tw/~cjlin/papers/guide/guide.pdf.. •‧ 國. ㈻㊫學. Hu, M., & Liu, B. 2004. Mining and summarizing customer reviews. Paper presented at the Proceedings of the tenth ACM SIGKDD international conference on Knowledge. •‧. discovery and data mining.. sit. y. Nat. Jonathan Milgram, Mohamed Cheriet, Robert Sabourin. 2006. One Against One” or ”One. al. er. io. Against All”: Which One is Better for Handwriting Recognition with SVMs?. Guy. n. Lorette. Tenth International Workshop on Frontiers in Handwriting Recognition, Oct 2006, La Baule. France. Ch. engchi. i Un. v. Ku, L.-W. & Chen, H.-H. 2007. Mining opinions from the web: beyond relevance retrieval. Journal of American Society for Information Science and Technology, 58. 12. ,. 1838-850. L.-W. Ku and H.-H. Chen. 2007. Mining opinions from the Web: Beyond relevance retrieval. Journal of the American Society for Information Science and Technology, vol. 58, no. 12, pp. 1838-1850. Lun-Wei Ku and Hsin-Hsi Chen. 2007. Mining opinions from the Web: Beyond relevance retrieval. Journal of the American Society for Information Science and Technology 58. Pang, Bo, Lillian Lee, and Shivakumar Vaithyanathan. 2002. Thumbs up? Sentiment !. 59!.
(69) !. Classification using Machine Learning Techniques. Empirical methods in natural language processing - Volume 10 Pages 79-86 Shiller, Robert J. 2007. Understanding Recent Trends in House Prices and Homeownership. NBER Working Paper No. 13553 Smith, B. A., and W. P. Tesarek. 1991. House Prices and Regional Real Estate Cycles: Market Adjustments in Houston. Journal of the American Real Estate and Urban Economics Association 19, 396-416. Su, C. W., and Lin, C. J. 2002. A Comparison of Methods for Multi-class Support Vector Machines, Proceedings of IEEE Transactions on Neural Networks, pp.415-425.. 政 治 大. T. Mullen and N. Collier 2004. Sentiment analysis using support vector machines with. 立立. diverse information sources. In Proceedings of Conference on Empirical Methods in. •‧ 國. ㈻㊫學. Natural Language Processing.. Vladimir Vapnik , Steven E. Golowich , Alex Smola. 1997. Support Vector Method for. •‧. Function Approximation, Regression Estimation, and Signal Processing. Neural. sit. n. al. er. io. !. y. Nat. Information Processing Systems Conference. Ch. engchi. 60!. i Un. v.
(70)
相關文件
(a) In your group, discuss what impact the social issues in Learning Activity 1 (and any other socials issues you can think of) have on the world, Hong Kong and you.. Choose the
another direction of world volume appears and resulting theory becomes (1+5)D Moreover, in this case, we can read the string coupling from the gauge field and this enables us to
If necessary, you might like to guide students to read over the notes and discuss the roles and language required of a chairperson or secretary to prepare them for the activity9.
Centre for Learning Sciences and Technologies (CLST) The Chinese University of Hong Kong..
FMEA, fail mode and effective analysis, which is one of a common method to analysis and find out the fail mode of the product is to dig out the unobservable problem or hidden
The research data are collected by questionnaire survey and analysed in electors’ opinion with political party of candidate, application of information technology media,
This study will use BIM, Web-base, RFID, Wireless Network, ER Model, Database and Information Technology environment to develop “Electronic Building Construction
Therefore, how to promote and the maintenance service quality can continue forever topic of the management, becomes the research once more focal point.So, how to try to
