GH-22: update report

undertheseanlp · Jan 14, 2019 · 12d3d93 · 12d3d93
1 parent 8018d0e
commit 12d3d93
Show file tree

Hide file tree

Showing 3 changed files with 17 additions and 65 deletions.
diff --git a/docs/technique_report.bib b/docs/technique_report.bib
@@ -1,64 +1,14 @@
-@article{DBLP:journals/corr/Le-Hong16,
-  author    = {Phuong Le{-}Hong},
-  title     = {Vietnamese Named Entity Recognition using Token Regular Expressions
-               and Bidirectional Inference},
-  journal   = {CoRR},
-  volume    = {abs/1610.05652},
-  year      = {2016},
-  url       = {http://arxiv.org/abs/1610.05652},
-  archivePrefix = {arXiv},
-  eprint    = {1610.05652},
-  timestamp = {Wed, 07 Jun 2017 14:42:34 +0200},
-  biburl    = {https://dblp.org/rec/bib/journals/corr/Le-Hong16},
-  bibsource = {dblp computer science bibliography, https://dblp.org}
-}
-
-@article{DBLP:journals/corr/abs-1708-07241,
-  author    = {Thai{-}Hoang Pham and
-               Xuan{-}Khoai Pham and
-               Tuan{-}Anh Nguyen and
-               Phuong Le{-}Hong},
-  title     = {{NNVLP:} {A} Neural Network-Based Vietnamese Language Processing Toolkit},
-  journal   = {CoRR},
-  volume    = {abs/1708.07241},
-  year      = {2017},
-  url       = {http://arxiv.org/abs/1708.07241},
-  archivePrefix = {arXiv},
-  eprint    = {1708.07241},
-  timestamp = {Tue, 05 Sep 2017 10:03:46 +0200},
-  biburl    = {https://dblp.org/rec/bib/journals/corr/abs-1708-07241},
-  bibsource = {dblp computer science bibliography, https://dblp.org}
-}
-
-@article{DBLP:journals/corr/abs-1801-01331,
-  author    = {Thanh Vu and
-               Dat Quoc Nguyen and
-               Dai Quoc Nguyen and
-               Mark Dras and
-               Mark Johnson},
-  title     = {VnCoreNLP: {A} Vietnamese Natural Language Processing Toolkit},
-  journal   = {CoRR},
-  volume    = {abs/1801.01331},
-  year      = {2018},
-  url       = {http://arxiv.org/abs/1801.01331},
-  archivePrefix = {arXiv},
-  eprint    = {1801.01331},
-  timestamp = {Thu, 01 Feb 2018 19:52:26 +0100},
-  biburl    = {https://dblp.org/rec/bib/journals/corr/abs-1801-01331},
-  bibsource = {dblp computer science bibliography, https://dblp.org}
-}
-
-@article{DBLP:journals/corr/abs-1803-08463,
-  author    = {Pham Quang Nhat Minh},
-  title     = {A Feature-Based Model for Nested Named-Entity Recognition at {VLSP-2018}
-               {NER} Evaluation Campaign},
-  journal   = {CoRR},
-  volume    = {abs/1803.08463},
-  year      = {2018},
-  url       = {http://arxiv.org/abs/1803.08463},
-  archivePrefix = {arXiv},
-  eprint    = {1803.08463},
-  timestamp = {Wed, 11 Apr 2018 11:12:46 +0200},
-  biburl    = {https://dblp.org/rec/bib/journals/corr/abs-1803-08463},
-  bibsource = {dblp computer science bibliography, https://dblp.org}
+@inproceedings{Lafferty:2001:CRF:645530.655813,
+ author = {Lafferty, John D. and McCallum, Andrew and Pereira, Fernando C. N.},
+ title = {Conditional Random Fields: Probabilistic Models for Segmenting and Labeling Sequence Data},
+ booktitle = {Proceedings of the Eighteenth International Conference on Machine Learning},
+ series = {ICML '01},
+ year = {2001},
+ isbn = {1-55860-778-1},
+ pages = {282--289},
+ numpages = {8},
+ url = {http://dl.acm.org/citation.cfm?id=645530.655813},
+ acmid = {655813},
+ publisher = {Morgan Kaufmann Publishers Inc.},
+ address = {San Francisco, CA, USA},
 }
diff --git a/docs/technique_report.pdf b/docs/technique_report.pdf
diff --git a/docs/technique_report.tex b/docs/technique_report.tex
@@ -47,9 +47,11 @@ \subsection{Hệ thống tách từ}
 
 \subsection{Thuật toán Conditional Random Fields}
 
-In order to solve word segmentation problem, there are many algorithms such as HMM, SVM, Riple Down Rules. In our experiments, we use condtional random fields, which yields many success for sequence labeling problem.
+Thuật toán Conditional Random Fields (CRFs) ~\cite{Lafferty:2001:CRF:645530.655813} được sử dụng đã tính toán xác suất của chuỗi đầu ra cho bởi chuỗi đầu vào. Xác suất của chuỗi trạng thái $S = <s_1, s_2,..., s_T>$ cho bởi quan sát $O = <o_1, o_2, ..., o_T>$ được tính bởi công thức:
 
-In this session, we brife describe conditional random fields algorithm.
+$$P(s|o) = \frac{1}{Z_o} exp( \sum_{t=1}^{T} \sum_{k} \lambda_k x f_k (s_{t-1},s_t,o,t) )$$
+
+trong đó, $f_k (s_{t-1},s_t,o,t)$ làm một hàm đặc trưng ứng với trọng số $\lambda_k$, được học thông qua quá trình huấn luyện.
 
 \subsection{Features}
 We propose conditional random fields for this problem.