TAILIEUCHUNG - Báo cáo khoa học: "An Algorithm for Unsupervised Transliteration Mining with an Application to Word Alignment"

We propose a language-independent method for the automatic extraction of transliteration pairs from parallel corpora. In contrast to previous work, our method uses no form of supervision, and does not require linguistically informed preprocessing. We conduct experiments on data sets from the NEWS 2010 shared task on transliteration mining and achieve an F-measure of up to 92%, outperforming most of the semi-supervised systems that were submitted. | An Algorithm for Unsupervised Transliteration Mining with an Application to Word Alignment Hassan Sajjad Alexander Fraser Helmut Schmid Institute for Natural Language Processing University of Stuttgart sajjad fraser schmid @ Abstract We propose a language-independent method for the automatic extraction of transliteration pairs from parallel corpora. In contrast to previous work our method uses no form of supervision and does not require linguistically informed preprocessing. We conduct experiments on data sets from the NEWS 2010 shared task on transliteration mining and achieve an F-measure of up to 92 outperforming most of the semi-supervised systems that were submitted. We also apply our method to English Hindi and English Arabic parallel corpora and compare the results with manually built gold standards which mark transliterated word pairs. Finally we integrate the transliteration module into the GIZA word aligner and evaluate it on two word alignment tasks achieving improvements in both precision and recall measured against gold standard word alignments. 1 Introduction Most previous methods for building transliteration systems were supervised requiring either handcrafted rules or a clean list of transliteration pairs both of which are expensive to create. Such resources are also not applicable to other language pairs. In this paper we show that it is possible to extract transliteration pairs from a parallel corpus using an unsupervised method. We first align a bilingual corpus at the word level using GIZA and create a list of word pairs containing a mix of nontransliterations and transliterations. We train a sta 430 tistical transliterator on the list of word pairs. We then filter out a few word pairs those which have the lowest transliteration probabilities according to the trained transliteration system which are likely to be non-transliterations. We retrain the translitera-tor on the filtered data set. This process is iterated filtering .

• Báo cáo khoa học
• An Algorithm
• Unsupervised Transliteration Mining
• Application to Word Alignment
• báo cáo khoa học
• báo cáo ngôn ngữ
• ngôn ngữ tự nhiên
• Solving uncapacitated multiple allocation p hub center problem
• Dijkstra’s algorithm based genetic algorithm and simulated annealing
• Dijkstra’s algorithm based genetic algorithm
• Large scale standard data sets
• P hub center problem
• Algorithm design
• Lecture Algorithm design
• Algorithm analysis
• Dynamic programming
• Hirschbergs algorithm
• Bellman Ford algorithm
• Computer Security
• Role Based Access Control
• Role Classification Algorithm
• Algorithm Preliminaries
• Algorithm Training Phase
• Algorithm Classification Phase
• Computational tractability
• Asymptotic order of growth
• Polynomial running time
• Greedy Algorithms
• Dijkstras algorithm
• Minimum spanning trees
• Network flow
• Ford Fulkerson algorithm
• Max flow min cut theorem
• International Journal of computer science & communication networks
• Anomaly detection using genetic
• SVM algorithm in data mining
• Algorithm in data mining
• K mean algorithm
• Two algorithm genetic and SVM
• Optimization of machining processes using pattern search algorithm
• Pattern search algorithm
• The PS algorithm
• Machining optimization problems
• Lecture Data Structure and Algorithm
• Bài giảng Cấu trúc dữ liệu và thuật toán
• Data Structure and Algorithm
• Sorting algorithm
• External sort
• Internal sort
• NTRU encryption algorithm
• Digital signature scheme
• Lectures on the NTRU encryption algorithm
• The NTRU encryption algorithm
• Some basic NTRU security issues
• Local search
• Gradient descent
• Metropolis algorithm
• Vietnam Journal of Earth Sciences
• Algorithm and program for earthquake prediction
• Algorithm and program based on the geological
• Algorithm and program based on the geophysical
• Algorithm and program based on the geomorphological
• The maximum magnitude values Mmax
• International Journal of Computer Networks and Communications Security
• New encryption algorithm
• Network RFWKIDEA8 1
• Transformation of AES encryption algorithm
• New block encryption algorithm
• Fault aware and congestion aware routing algorithm
• On Odd Even algorithm
• Network on chip
• The Fault Aware Dynamic Routing algorithm
• Advanced Algorithms Analysis and Design
• Lecture Advanced Algorithms Analysis and Design
• Bài giảng Phân tích và thiết kế thuật toán nâng cao
• The Floyd Warshall algorithm
• Johnson’s algorithm
• A Hardware implementation
• Winograd Fourier Transform
• Discrete fourier transform
• Winograd fourier transform algorithm
• Winograd’s small convolution algorithm
• Transform algorithm for Cryptography
• Graph search
• Graph connectivity
• Graph traversal
• Coin changing
• Optimal caching
• Counting inversions
• Closest pair of points
• Randomized quicksort
• Master theorem
• Integer multiplication
• Matrix multiplication
• Weighted interval scheduling
• Knapsack problem
• Bipartite matching
• Disjoint paths
• Assignment problem
• Input queued switching
• Poly time reductions
• Constraint satisfaction problems
• Graph coloring
• Decision problems
• NP complete
• Nondeterministic polynomial