200/4/28,12:08 A DeepLEX: yet another Lexicon? Page 1 of 35 DeepLEX: yet another Lexicon? Shu-Kai Hsieh. Yu-Hsiang Tseng. Yong-Fu Liao. Han-Tang Hong - Graduate Institute of Linguistics, National Taiwan University -
DeepLEX: yet another Lexicon?Page 1 of 35
DeepLEX: yet another Lexicon? Shu-Kai Hsieh. Yu-Hsiang Tseng.
Yong-Fu Liao. Han-Tang Hong - Graduate Institute of Linguistics,
National Taiwan University -
DeepLEX: yet another Lexicon?
Page 2 of 35
What is an ideal Lexicon Model for (NLP/NLU)?-
DeepLEX: yet another Lexicon?
Page 4 of 35
Language Complexity in Communicative Settings
visual, auditory, verbal, and invisible suprasystems interacting in
a congruent state“ Polymorphic behaviour: A sign (word, phrase) can
have multiple senses in varied contexts. Resulting ambiguity in
use: lexical, structural, scope ( ,() ), anaphora/deictic
expressions etc.
Paralinguistic and extralinguistic information merging in
communicative settings: kinesics (body gestures), acoustic (vocal
activity), silence, etc.
·
·
Affective Expressions
Affect is a broader all-encompassing term whch refers to general
topics of emotion, feelings, and mood together.
Emotion, Mood, Feeling, Attitude, Temperament, Personal
trait·
5/35
Expressive Units of Emotions Affect Lexicon oversimplifies the
affaction process
8/35
·
·
Reading (evaluative) texts
·
Multimodality affect state (paralinguistics)
Sense-sentiment concurrent processing
·
·
$#* :-(“
tokens(" $#*")
## tokens from 1 document. ## text1 : ## [1] "" "" "" "" "" "" ""
## [8] "" "" "" "" "" "" "" ## [15] "" "" "" "" "" "" "" "*" "#"
"$" "" "" [22] ##
·
·
·
DeepLEX: yet another Lexicon?
Page 17 of 35
Wordnet·
-
19/35
Meanings are negotiated, subordinated to the sequential
requirements of talk-in-interaction.
·
·
DeepLEX: yet another Lexicon?
Page 22 of 35
Pre-packageed information (formulaic expressions) relatively immune
from negotiation, which figure prominently in oral discourse, and
significantly often coincide with the boundaries of intonational
units, where syntactic and pragmatic completion points often
converge.(Huang, 1995).“
22/35
DeepLEX: Assumptions
It takes the functional position (usage-based view) in determining
units and patterns (in Chinese), as well as the ontological
grounding on the relation between linguistic objects and situations
(bits of reality). (Langacker 1987, 1988, 1999; Croft 2002;
Tomasello 2003; Bybee 2006, 2010)
·
·
library(jiebaR) seg <- worker() seg[" der"]
## [1] "" "" "" "" "" "" "" ## [8] "" "" "" "" "der"
24/35
txt_khmer <-
"!#$&'()+,'-.)/034678i;=/>?AB47DB3&FH?JK47Jei " #Taiwan
Steps up Asia Business to Reduce Dependence on China #taivean
baohchomhan chhpaohtow rokkarothveu peanechchokamm now asai daembi
#katbanthoy pheap asry leu bratesa chen tokens(txt_khmer)
## tokens from 1 document. ## text1 : ## [1] "!#$&" "'()+,"
"'-.)/0" "3467" "8i;=" "/>" "?AB" ## [8] "47DB" "3&F" "H"
"?JK" "47" "Je" "i"
25/35
Character variants ()
·
·
·
·
·
phonetics sense polarity 1930.freq 3y.freq indegree POS —–
components relations classes 1940.freq 4y.freq outdegree
productivity —–
—— —— —– —– —– —– —– —–
“
DeepLEX-based pilot sudies:
Affective dialogue system
Mega- and meta-analysis
## Loading `database.rds`...
Huang, S.F. 1995. Emergent Lexical Semantics.
Hsieh, S.K. et al. 2019. Fluid Annotation: A Granularity-aware
Annotation Tool for Chinese Word Fluidity. LREC. 2019.
Hsieh, Shu-Kai and Yu-Hsiang Tseng. 2019. Linguistic Granularity
Annottion Framework: a granularity-aware approach to Chinese NLP.
Journal of Granular Computing. Springer. (under review).
Tseng, Yu-Hiang and Shu-Kai Hsieh. 2019. Eigencharacters.
2019. : 61(3).