Final-modal particles 語氣詞 Final-modal particles 褚 靚 A03142104
Key dealing corpus annotate analyze
Corpus/ Lexicon Build personal Corpus Lexicon
Corpus 臺灣政治大學漢語口語語料庫 NCCU Corpus of Spoken Chinese 聯合語料庫-聯合報 ? PTT ? 聯合語料庫-聯合報 ? 中國傳媒大學媒體語言語料庫
PASS Summit 2011 11/15/2018 © 2011 Microsoft Corporation. All rights reserved. Microsoft, Windows, Windows Vista and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.
Lexicon Lexicon台 Lexicon陸 各自框架-lexicon 共同高頻-corpus
WHO/HOW 台生 台語料 陸生 陸語料 Annotate PASS Summit 2011 11/15/2018 © 2011 Microsoft Corporation. All rights reserved. Microsoft, Windows, Windows Vista and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.
HOW? Annotate Sentence Polarity Confirmative degree Chunk based End/ Beginning Gender tendency S1:————□1 N S2:————□1 S3:————□1 S4:————□1 Y S5:————□1 S6:————□1 S7:————□1 HOW?
Average the different ideas Counting Predictive polarity Valence Calculate the average emotion polarity valence of each keyword pos=1, neg=-1,neu=0 X+Y+Z=N (A constant value=all people) X = The number of people prefer this sentence positive polarity Y = The number of people prefer this sentence negative polarity Z = The number of people prefer this sentence neutral polarity X+Y+Z= Total number of the people participating in annotation work Predictive_ Polarity S1 = X*1+y*(-1) N Average the different ideas
Average whole sentences Counting Predictive polarity Valence Calculate the average emotion polarity valence of each keyword pos=1, neg=-1,neu=0 S1+S2+……+Sn-1+Sn=M (A constant value=all sentences) Polarity_S1 (P1)= -0.3 Polarity_S2 (P2)= -0.6 ………………………………………… Polarity_Sn (Pn)= 0.2 S1+S2+……+Sn-1+Sn= Total number of the sentences being annotated Predictive_ Polarity Valence = P1+S2+……+Pn M Average whole sentences
An example Analyze
An example in comparison between”嗎”and”嘛” PASS Summit 2011 11/15/2018 An example in comparison between”嗎”and”嘛” © 2011 Microsoft Corporation. All rights reserved. Microsoft, Windows, Windows Vista and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.
Ideas welcomed!
Thanks for listening 2015.5.21