taegyeonglee commited on
Commit
10154c7
ยท
verified ยท
1 Parent(s): a6e6817

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +106 -1
README.md CHANGED
@@ -5,4 +5,109 @@ language:
5
  - en
6
  base_model:
7
  - klue/bert-base
8
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
5
  - en
6
  base_model:
7
  - klue/bert-base
8
+ ---
9
+ # LQ-KBERT-Base: Crypto Market Korean Sentiment & Action Signal Classifier
10
+
11
+ [LangQuant](https://langquant.com)์—์„œ ๊ณต๊ฐœํ•œ **ํ•œ๊ตญ์–ด ๊ธˆ์œต ์ปค๋ฎค๋‹ˆํ‹ฐ/๋‰ด์Šค ํˆฌ์ž์‹ฌ๋ฆฌ ๋ถ„๋ฅ˜ ๋ชจ๋ธ**์ž…๋‹ˆ๋‹ค.
12
+ `klue/bert-base`๋ฅผ ๋ฐฑ๋ณธ์œผ๋กœ ํ•˜๊ณ , ๊ฐ€์ƒ์ž์‚ฐ ๊ด€๋ จ ํ•œ๊ตญ์–ด ๋ฐ์ดํ„ฐ์…‹ **10๋งŒ ๊ฑด ์ด์ƒ**์„ ์ „์ฒ˜๋ฆฌํ•˜์—ฌ ํŒŒ์ธํŠœ๋‹ํ–ˆ์Šต๋‹ˆ๋‹ค.
13
+ ๋ชจ๋ธ์€ ๋ฌธ์žฅ ๋‹จ์œ„ ์ž…๋ ฅ(`โ‰ค200์ž`)์— ๋Œ€ํ•ด **ํˆฌ์ž ์‹ฌ๋ฆฌยทํ–‰๋™ยท๊ฐ์ •ยทํ™•์‹ ๋„ยท๊ด€๋ จ์„ฑยท์œ ํ•ด์„ฑ**์„ ๋™์‹œ์— ์˜ˆ์ธกํ•ฉ๋‹ˆ๋‹ค.
14
+
15
+ - [Github](https://github.com/LangQuant/LQ-KBERT-Base)
16
+ ---
17
+ ### ๋ชจ๋ธ์€ ๋‹ค์Œ ํ•ญ๋ชฉ์„ ์˜ˆ์ธกํ•ฉ๋‹ˆ๋‹ค.
18
+
19
+ ```json
20
+ {
21
+ "sentiment_strength": "strong_pos | weak_pos | neutral | weak_neg | strong_neg",
22
+ "action_signal": "buy | hold | sell | avoid | info_only | ask_info",
23
+ "emotions": ["greed","fear","confidence","doubt","anger","hope","sarcasm"],
24
+ "certainty": 0.0 ~ 1.0,
25
+ "relevance": 0.0 ~ 1.0,
26
+ "reasons": "๋ผ๋ฒจ ๊ทผ๊ฑฐ๋ฅผ ์š”์•ฝํ•œ ํ•œ๊ตญ์–ด 1~2๋ฌธ์žฅ",
27
+ "toxicity": 0.0 ~ 1.0
28
+ }
29
+ ```
30
+ ---
31
+ ## Labeling Guidelines
32
+
33
+ ### Sentiment Strength
34
+ - **strong_pos**: ๊ธ‰๋“ฑ ํ™•์‹ , `"๊ฐ€์ฆˆ์•„"`, `"๋ฌด์กฐ๊ฑด ๊ฐ„๋‹ค"`.
35
+ - **weak_pos**: ์กฐ์‹ฌ์Šค๋Ÿฌ์šด ๋‚™๊ด€, `"๋ฐ˜๋“ฑ ๊ฐ€๋Šฅ"`, `"๊ดœ์ฐฎ์„ ๋“ฏ"`.
36
+ - **neutral**: ๋‹จ์ˆœ ์ •๋ณด/๊ณต์ง€/์žก๋‹ด.
37
+ - **weak_neg**: ์™„๊ณกํ•œ ๋ถ€์ •, `"์กฐ์ • ์˜ฌ ๋“ฏ"`, `"๊ด€๋ง"`.
38
+ - **strong_neg**: ํญ๋ฝยทํŒจ๋‹‰, `"๋‚˜๋ฝ"`, `"๋งํ•จ"`, `"ํ•ดํ‚น/์ œ์žฌ"`.
39
+
40
+ ### Action Signal
41
+ - **buy**: ๋งค์ˆ˜/์ง„์ž… ์ง€์‹œ, `"์ง€๊ธˆ ์‚ฐ๋‹ค"`, `"๋กฑ"`.
42
+ - **hold**: ๋ณด์œ  ์œ ์ง€/๊ด€๋ง, `"์กด๋ฒ„"`, `"์œ ์ง€"`.
43
+ - **sell**: ๋งค๋„/์ฒญ์‚ฐ, `"์ต์ ˆ"`, `"์†์ ˆ"`, `"์ •๋ฆฌ"`.
44
+ - **avoid**: ํšŒํ”ผ/์œ„ํ—˜ ๊ฒฝ๊ณ , `"๊ฐ€์ง€๋งˆ"`, `"์Šค์บ "`, `"์œ„ํ—˜"`.
45
+ - **info_only**: ๋‹จ์ˆœ ์ •๋ณด ์ „๋‹ฌ (๋‰ด์Šค/๊ณต์ง€).
46
+ - **ask_info**: ์งˆ๋ฌธ/ํƒ์ƒ‰, `"๋“ค์–ด๊ฐ€๋„ ๋ผ?"`, `"์™œ ๋–จ์–ด์ ธ?"`.
47
+
48
+ ### Emotions (๋‹ค์ค‘ ์„ ํƒ)
49
+ - **greed** ํƒ์š•
50
+ - **fear** ๋‘๋ ค์›€
51
+ - **confidence** ํ™•์‹ 
52
+ - **doubt** ์˜์‹ฌ
53
+ - **anger** ๋ถ„๋…ธ
54
+ - **hope** ํฌ๋ง
55
+ - **sarcasm** ํ’์ž
56
+
57
+ ### Certainty
58
+ - **0.2~0.4**: ์งˆ๋ฌธยทํƒ์ƒ‰ยท๋ฐˆ (๋‚ฎ์Œ)
59
+ - **0.4~0.6**: ์™„๊ณกํ•œ ์˜๊ฒฌ (์ค‘๊ฐ„)
60
+ - **0.6~0.8**: ์ˆ˜์น˜ยท๊ทผ๊ฑฐยท๊ณต์‹์„ฑ (๋†’์Œ)
61
+ - **0.8~1.0**: ๊ฐ•ํ•œ ๋‹จ์ •ยท์ง€์‹œ (๋งค์šฐ ๋†’์Œ)
62
+
63
+ ### Relevance
64
+ - **0.7~1.0**: ์ง์ ‘์ ์ธ ํˆฌ์ž/์‹œ์žฅ ๊ด€๋ จ
65
+ - **0.4~0.7**: ๊ฐ„์ ‘ ๊ด€๋ จ (์—…๊ณ„/์ธ๋ฌผ/๊ธฐ์ˆ )
66
+ - **0.0~0.3**: ๋ฌด๊ด€/์žก๋‹ด/๋ฐˆ
67
+
68
+ ### Toxicity
69
+ - ์š•์„คยท๋ชจ์š•ยท๋น„ํ•˜ ๊ฐ•๋„์— ๋”ฐ๋ผ **0~1**.
70
+ - ํˆฌ์ž ์˜๋ฏธ์™€๋Š” ๋ณ„๋„๋กœ ๋…๋ฆฝ์ ์œผ๋กœ ํ‰๊ฐ€.
71
+
72
+ ---
73
+
74
+ ## Sentiment Strength vs Action Signal
75
+
76
+ - **Sentiment Strength**
77
+ - ํˆฌ์ž ์‹ฌ๋ฆฌ์˜ ๊ฐ•๋„ (๊ธ์ • โ†” ๋ถ€์ •).
78
+ - ๊ฐ€๊ฒฉ ์ „๋ง์˜ ํ†ค์— ์ง‘์ค‘.
79
+
80
+ - **Action Signal**
81
+ - ์‹ค์ œ ํˆฌ์ž ํ–‰๋™ ์˜๋„/์ง€์‹œ.
82
+ - ๋งค์ˆ˜/๋งค๋„/๋ณด์œ /ํšŒํ”ผ/์งˆ๋ฌธ/์ •๋ณด.
83
+
84
+
85
+ ---
86
+
87
+ ### ์˜ˆ์‹œ
88
+
89
+ | ๋ฌธ์žฅ | sentiment_strength | action_signal | ํ•ด์„ |
90
+ |------|--------------------|---------------|------|
91
+ | "๊ฐœ๋–ก์ƒ์ด์—ฌ " | strong_pos | buy | ๊ฐ•ํ•œ ์ƒ์Šน ํ™•์‹  + ์ฆ‰์‹œ ๋งค์ˆ˜ ์˜๋„ |
92
+ | "์—ฌ๊ธฐ์„  ๊ด€๋ง์ด ๋งž๋‹ค" | weak_neg | hold | ๋ถ€์ •์ ์ด์ง€๋งŒ ๋ณด์œ  ์œ ์ง€ ์„ ํƒ |
93
+ | "๋“ค์–ด๊ฐ€๋„ ๋ ๊นŒ?" | weak_pos | ask_info | ์กฐ์‹ฌ์Šค๋Ÿฌ์šด ๋‚™๊ด€, ๋งค์ˆ˜ ํƒ์ƒ‰ ์งˆ๋ฌธ |
94
+ | "ํ•ดํ‚น ํ„ฐ์ง, ๋น„์ƒ. ์ ‘๊ทผ ๊ธˆ์ง€" | strong_neg | avoid | ๊ฐ•ํ•œ ๋ถ€์ • + ํšŒํ”ผ ๊ถŒ๊ณ  |
95
+ | "์—…๋ฐ์ดํŠธ ๊ณต์ง€ ๋‚˜์™”์Šต๋‹ˆ๋‹ค" | neutral | info_only | ๋‹จ์ˆœ ์ •๋ณด ์ œ๊ณต, ํ–‰๋™ ์—†์Œ |
96
+
97
+ ---
98
+ ### Citation
99
+ ```
100
+ @misc{langquant2025lkbert,
101
+ title = {LQ-KBERT-Base: Crypto Market Korean Sentiment & Action Signal Classifier},
102
+ author = {LangQuant},
103
+ year = {2025},
104
+ url = {https://huggingface.co/langquant/LQ-Kbert-base}
105
+ }
106
+ ```
107
+ ---
108
+ ### Disclaimer
109
+ ```
110
+ ์ด ๋ชจ๋ธ์€ ํ•™์ˆ  ์—ฐ๊ตฌ ๋ฐ ์‹คํ—˜์šฉ์œผ๋กœ๋งŒ ์ œ๊ณต๋ฉ๋‹ˆ๋‹ค.
111
+ ๋ณธ ๋ชจ๋ธ์˜ ์ถœ๋ ฅ์€ ๊ธˆ์œต/ํˆฌ์ž ์ž๋ฌธ์œผ๋กœ ๊ฐ„์ฃผ๋  ์ˆ˜ ์—†์œผ๋ฉฐ,
112
+ ๋ฐœ์ƒํ•˜๋Š” ๋ชจ๋“  ๊ฒฐ๊ณผ์— ๋Œ€ํ•ด LangQuant๋Š” ์ฑ…์ž„์„ ์ง€์ง€ ์•Š์Šต๋‹ˆ๋‹ค.
113
+ ```