不自然言語処理コンテストLT資料
2 likes3,109 views
This document discusses emoji analysis and frequency counting. It describes using a Double Array Trie Tree structure called Darts to efficiently store and search emoji sequences and their frequencies from a large corpus. Examples are given of frequent emoji sequences extracted from Twitter along with their counts. Links are provided to resources about the Unicode emoji standard and the Darts algorithm.
![→ →
→ http://www.unicode.org/~scherer/emoji4unicode/snapshot/full.html
※[ 2010]
http://nlp.nagaokaut.ac.jp/arc/10/10NLP-asakura.pdf](https://image.slidesharecdn.com/20100725baidu-100725104632-phpapp02/85/LT-11-320.jpg)
![["EMOJI_B36","EMOJI_B60","EMOJI_000","EMOJI_B95"] 393
["EMOJI_B05"] 80
) ["EMOJI_B85"] 160
nexi■ ["EMOJI_B05"] 450
["EMOJI_B85"] 365
["EMOJI_B05"] 136
["EMOJI_B05"] 686
["EMOJI_B60"] 67
["EMOJI_525"] 3330
※](https://image.slidesharecdn.com/20100725baidu-100725104632-phpapp02/85/LT-12-320.jpg)
不自然言語処理コンテストLT資料
- 1. n 7/25 at Baidu Office http://twitter.com/just_do_neet
- 2. ‣ http://twitter.com/just_do_neet ‣ http://ameblo.jp/just-do-neet/
- 3. - -
- 5. Baidu
- 6. Baidu n - -
- 8. UI
- 9. - n • • UI
- 10. n <EMOJI_xxx> <EMOJI_000> 48 → <S> <EMOJI_000> 101 →
- 11. → → → http://www.unicode.org/~scherer/emoji4unicode/snapshot/full.html ※[ 2010] http://nlp.nagaokaut.ac.jp/arc/10/10NLP-asakura.pdf
- 12. ["EMOJI_B36","EMOJI_B60","EMOJI_000","EMOJI_B95"] 393 ["EMOJI_B05"] 80 ) ["EMOJI_B85"] 160 nexi■ ["EMOJI_B05"] 450 ["EMOJI_B85"] 365 ["EMOJI_B05"] 136 ["EMOJI_B05"] 686 ["EMOJI_B60"] 67 ["EMOJI_525"] 3330 ※
- 13. - Trie - Darts Double Array Trie Tree • Common Prefix Search UI
- 14. Darts ‣ http://chasen.org/~taku/software/darts/
- 15. UI - • JavaScript(JQuery • Java
- 17. -