ROCLING_手機新聞摘要.doc

Size: px

Start display at page:

Download "ROCLING_手機新聞摘要.doc"

互保章
7 years ago
Views:

1 % 65% 75%80% 20%25% [1] 100% [2] [13] PChome [4] [5] [6] 158 letter [2] PHS 45 [7-8] information overload IBM

2 [9]InXight [10]Megaputer [11] informative indicative commentary extraction abstraction intrinsic extrinsic coverage readability coherence cohesion organization conjunction pronoun anaphor rhetoric SUMMUC [12] DUC [8] NTCIR TSC [13] DUC words DUC [7]DUC NTCIR TSC 2002 Mainichi 20%40% [13] [14-18] 10% 10%-50% 10% clustering 90 76% 70% [19] Global Bushy PathGBP [20] GBP 89.77% 58.25% Banko [21] Kennedy [22] Corston-Oliver Monday Mon MicroSoft Outlook [23] Buyukkokten [24] Yang [25] PDA

3 45 real-time 30 [26] meaningful unit, MU FTTB L2 Switch /08/14, 11:21:28 Ethernet- based FTTB L2 Switch GESWr EoVDSL SMC /09/01, 07:57: MU

4 n w Keywords ( * tf / max_ tf ) w tf w w max_tf Tseng maximally repeated string[27] 12 n dynamic programming edit distance [28] AB nm A i A[i] B j B[j] d[i, j] = min( d[i-1, j] + w(a[i], 0), d[i-1, j-1] + w(a[i], B[j]), d[i, j-1] + w(0, B[j]) ) min(x, Y, Z) X, Y, Z d[0, 0] = 0 d[i, 0] = d[i-1, 0] + w(a[i], 0), 1<= i <= n d[0, j] = d[0, j-1] + w(0, B[j]), 1<= j <= m w(x, Y) w(a[i], B[j]) : A[i] B[j] w(a[i], 0) : A[i] w(0, B[j]) : B[j] A=adc B=adecdecf 1 d[i, j] i j B A a d e c d e c f A D C

5 B adec A B adec A adc adcdecf A=adcB=decadecf A B d e c a d e c f a d c [28] d[0, 0] = 0 d[i, 0] = d[i-1, 0] + w(a[i], 0), 1<= i <= n d[0, j] = 0, 1<= j <= m A B d e c a d e c F a d c Lopresti [28] exp(edit/(edit-m)) natural exponentedit m m m+n n edit sim = exp( ) = edit edit m n 1 n+ m edit e 1 1/exp(m/n) k k n A B B A A A B 1.25 A B 0.75 A B

6 G F B 45 1, =0.8767, =0.9333, 42 2, =0.8636, =0.8000, 36 1, =0.8636, =0.9130, 63 2, =0.8636, =0.5217, * Title Title 5 45 Title G Title B Title B 57 Title Title Title B 54 Title Title

7 Title G Title G 50 4 G Title Title F 68 Title B 64 Bermuda Global Fund Services Limited GFS Title ARM 3 45 ARM ARM ARM1026EJ -S(TM) (SOC) Title Inno Micro nstor 2 31 Inno Micro nstor nstor 1 B 54 Inno Micro nstor Inno Micro nstor Title Title Title Title G 63 Title Title Title G Title Title Title FTTBL2Switch 2 39 FTTBL2Switch 58 FTTBL2Switch SMC

8 Title 2500 (2-1) (2-1)4C 2500 (2-1) 4C Title (2-2) 3 41 (2-2) 2 G 63 (2-2) Title Title G G Title 29 ADR ADR % ADR B 0.68% Title B Title G Title Title Title Title Title G Title B 15 1 B 62.5%65% 75% 80% 20%25% (65.0%) 2 (5.0%) 5 (12.5%) 1 25 (62.5%) 6 (15%) 4 (10%) 2 5 (12.5%) 1 (2.5%) (7.5%) (2.5%) (2.5%) (2.5%) (80.0%) 3 (7.5%) 5 (12.5%) 30 (75%) 6 (15%) 4 (10%)

9 /5 1/ NSC E [1] ETtoday.com 2002/08/09 accessed on 2003/12/3. [2] egolife on 2003/12/3. [3] on 2003/12/3. [4] on 2003/12/3. [5] accessed on 2003/12/3. [6] by_author.asp?mmcontentnoid=43accessed on 2003/12/3. [7] Chin-Yew Lin and E.H. Hovy, The Potential and Limitations of Sentence Extraction for Summarization, Proceedings of the Workshop on Automatic Summarization, post-conference workshop of HLT-NAACL-2003, Edmonton, Canada, May 31 - June 1, [8] The Document Understanding Conference, [9] IBM, Intelligent Miner for Text: Summarization Tool, www-3.ibm.com/software/data/iminer/ fortext/summarize/summarize.html. [10] InXight,

10 [11] A List of Summarization Projects, accessed on 2003/12/3. [12] The TIPSTER SUMMAC Text Summarization Evaluation : Final Report, Oct., 1999, accessed on 2003/12/3 [13] Takahiro Fukusima, Manabu Okumura, and Hidetsugu Nanba, Text Summarization Challenge 2: Text Summarization Evaluation at NTCIR Workshop3, Proceedings of the Third NTCIR Workshop on Evaluation of Information Retrieval, Automatic Text Summarization and Question Answering, Oct. 8-10, 2002, Tokyo, Japan, pp.1-6. [14] : 2000 NSC E [15] (II)--- IV NSC E [16] (III)--- IV NSC E [17] Hsin-Hsi Chen, June-Jei Kuo, Tsei-Chun Su, Clustering and Visualization in a Multi-lingual Multi-document Summarization System, Proceedings of the 25th European Conference on IR Research, ECIR 2003, Pisa, Italy, April 14-16, 2003, pp [18] June-Jei Kuo, Hung-Chia Wung, Chuan-Jie Lin, Hsin-Hsi Chen, Multi-document Summarization Using Informative Words and Its Evaluation with a QA System, Proceedings of the Third International Conference Computational Linguistics and Intelligent Text Processing, Mexico City, Mexico, February 17-23, 2002, pp [19] (II) 2001 NSC E [20] ( ) 2001 NSC H [21] Michele Banko, Vibhu O. Mittal, Michael J. Witbrock, Headline Generation Based on Statistical Translation, Proceedings of the ACL [22] Paul E. Kennedy, Alexander G. Hauptmann, Automatic title generation for EM, Proceedings of the fifth ACM conference on Digital libraries, 2000, San Antonio, Texas, U.S., pp [23] Simon Corston-Oliver, Text Compaction for Display on Very Small Screens, In Proceedings of the Workshop on Automatic Summarization (WAS 2001), Pittsburgh, PA, USA. [24] O. Buyukkokten, H. Garcia -Molina, A. Paepcke Seeing the Whole in Parts: Text Summarization for Web Browsing on Handheld Devices, The 10th International WWW Conference (WWW10). Hong Kong, China. [25] Christopher C. Yang, Fu Lee Wang, Adapting content to mobile devices: Fractal summarization for mobile devices to access large documents on the web, Proceedings of the twelfth international conference on World Wide Web, May 2003, pp [26] [27] Yuen-Hsien Tseng, "Automatic Thesaurus Generation for Chinese Documents", Journal of the American Society for Information Science and Technology, Vol. 53, No. 13, 2002, pp [28] Daniel Lopresti and Jiangying Zhou, Retrieval Strategies for Noisy Text, Proceedings of the Fifth Annual Symposium on Document Analysis and Information Retrieval, April 15-17, 1996, pp <?xml version="1.0" encoding="big5"?> <!DOCTYPE Summary SYSTEM "guideline.dtd"> <News> ID, Title, Body 1.Title 2.Body <ID>chd_eco_ _0001</ID> <Title>

11 </Title> <Body> </Body> <Analysis> Analysis proper name 5W1H proper names <ProperName> 1. ProperName tag tag 4. tag tag tag tag 8. ( ) tag tag <person title=" "> </person> <person title=" "> </person> <person title=" "> </person> <organization abbreviation=""> </organization> <organization abbreviation=""> </organization> <organization abbreviation=" "> </organization> <organization abbreviation=""> </organization> <organization abbreviation=""> </organization> <location abbreviation=""> </location> </ProperName> <who> who ProperName ProperName 4. <name> </name> </who> <what> what

12 4. 5. <event> </event> <event> </event> <event> </event> <event> </event> </what> <when> when <date>1999/01/11</date> </when> <where> where ProperName 1. ProperName ProperName 2. <place> </place> <place> </place> <place> </place> </where> <why> why <reason> </reason> </why> <how> how <action> </action> <action> </action> <action> </action> <action> </action> </how> <HumanSummary CharLimit="" char="67"> 1. CharLimit="" char word </HumanSummary> <HumanExtract1 CharLimit="" char="63" CharModified="0">

13 1. tag 2. CharModified=" " CharModified="0" 3. char word 4. </HumanExtract1> <HumanExtract2 CharLimit="" char="63"> </HumanExtract2> </Analysis> </News> DRAM DRAM DRAM 64MDRAM DRAM DRAM 64MDRAM DRAM 64MDRAM 64MDRAM 16MDRAM 64MDRAM <HumanExtract2 CharLimit="" char="61"> </HumanExtract2> 1, =0.8304, =0.9420, 65 2, =0.258, =0.9420, 65 3, =0.7146, =0.86, 60 64MDRAM DRAM 4, =0.18, =0.57, 48 5, =0.4961, =0.7391, 51

第一章

第一章 6.1 1. N-gram 2. 3. 4. 60 6.2 4.114.13 4.124.14 Rong Jin[7] TF*IDF - SARS SARS SARS SARS 61 62 [1] Michele Banko, Vibhu O. Mittal, and Michael J. Witbrock. 2000. Headline Generation Based on Statistical