Báo cáo khoa học: "Using Deep Morphology to Improve Automatic Error Detection in Arabic Handwriting Recognition" pot

... Effect of Training Data Size In order to allow for rapid examination of multi- ple feature combinations, we restricted the size of the training set (S) to maintain manageable train- ing times. ... Computational Linguistics Using Deep Morphology to Improve Automatic Error Detection in Arabic Handwriting Recognition Nizar Habash and Ryan M. Roth Center for Computationa...

Ngày tải lên: 17/03/2014, 00:20

... was set to. maxlinks Used in Competitive Linking algo- rithm: Maximum number of words any word can be aligned with. Set to: 1, 2, 3. minscore Used in Competitive Linking algo- rithm: Minimum score ... is 2, then a word can align to 0, 1, or 2 words in the parallel sentence. In other settings, we en- forced a minimum score in the bilingual dictionary for a link to be accepted,...

Ngày tải lên: 08/03/2014, 07:20

... effective, while being simpler to instrument: since we use information from the search engine only during training, we can train a stand-alone POS tagger that can be run without access to additional resources. ... query training set. The simplest way to match query tokens to snip- pet tokens is to allow a query token to match any snippet token. This can be problematic when we have...

Ngày tải lên: 16/03/2014, 20:20

... Darmstadt http://www.ukp.tu-darmstadt.de/people Abstract Current work on automatic opinion min- ing has ignored opinion targets expressed by anaphorical pronouns, thereby missing a significant number of opinion targets. In this paper we empirically ... This aboutness has been referred to as the opinion tar- get or opinion topic in the literature from the field. In this work our goal i...

Ngày tải lên: 17/03/2014, 00:20

... protein may be necessary to maintain the structure of the C-terminal domain in a conformation that allows the metals to bind. Removing the N-terminal domain results in a C-terminal domain in which ... single-domain AMPM protein [3]. In the case of AMPP, the N-terminal domain appears to have a function in protein folding. Clearly, the single- domain AMPM protein has found anoth...

Ngày tải lên: 23/03/2014, 07:20

... implementing these analyses. In this paper we present a method of identifying prob- lems in a deep grammar by exploiting the fact that it can be used for both parsing (interpreting text into semantics) ... Sometimes the 1 Deep Linguistic Processing with HPSG Initiative – see http://www.delph -in. net for background informa- tion, including the list of current participants and pointers...

Ngày tải lên: 31/03/2014, 00:20

... boundaries are missing. Sentence segmentation information is crucial and as- sumed in most of the further processing steps that one would want to apply to such output: tagging and parsing, information ... parameters. Since the training set for the SU detection task in the EARS program is quite limited, we use a loosely coupled approach: Linearly combine three LMs: the word-based LM...

Ngày tải lên: 31/03/2014, 03:20

... as domain independent pivots and enable us to transfer the information regarding sentiment from one domain to another. Using the extended vectors d  to represent re- views, we train a binary ... cross-domain sentiment classification. In our method, we use the automatically created thesaurus to expand feature vectors in a binary clas- sifier at train and test times by introducing re...

Ngày tải lên: 20/02/2014, 04:20

... grammar, in which the subtrees are specified in advance, in an adaptor grammar the subtrees, as well as their probabilities, are learnt from the train- ing data. In order to make parsing and inference tractable ... way to capture dependencies between words in collocations except to make the collocation into a single word. 3.2 Unigram morphology adaptor grammar This section in...

Ngày tải lên: 20/02/2014, 09:20

... try to insert *U* in X 7 try to insert a VP ellipsis site in X 8 try to insert S*T* or SBAR in X 9 try to insert trace of topicalized XP in X 10 try to insert trace of extraposition in X ... applying a different set of rules. 1 for each tree, iterate over nodes from top down 2 for each node X 3 try to insert NP* in X 4 try to insert 0 in X 5 try to ins...

Ngày tải lên: 20/02/2014, 16:20