• ETRI Journal
• /
• v.25 no.4
• /
• pp.219-230
• /
• 2003

This paper describes a rule-based approach for syntactic disambiguation used by the English sentence parser in E-TRAN 2001, an English-Korean machine translation system. We propose Parser's Ambiguity Type Information (PATI) to automatically identify the types of ambiguities observed in competing candidate trees produced by the parser and synthesize the types into a formal representation. PATI provides an efficient way of encoding knowledge into grammar rules and calculating rule preference scores from a relatively small training corpus. In the overall scoring scheme for sorting the candidate trees, the rule preference scores are combined with other preference functions that are based on statistical information. We compare the enhanced grammar with the initial one in terms of the amount of ambiguity. The experimental results show that the rule preference scores could significantly increase the accuracy of ambiguity resolution.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1995.10a
• /
• pp.124-140
• /
• 1995

The merit of fuzzy rule based systems stems from their capability of encoding qualitative knowledge of experts into quantitative rules. Recent advancement in automatic tuning or self-organization of fuzzy rules from experimental data further enhances their power, allowing the integration of the top-down encoding of knowledge with the bottom-up learning of rules. In this paper, methods of self-organizing fuzzy rules and of performing defuzzification and inference is presented based on a multi-resolution radial basis function network. The network learns an arbitrary input-output mapping from sample distribution as the union of hyper-ellipsoidal clusters of various locations, sizes and shapes. The hyper-ellipsoidal clusters, representing fuzzy rules, are self-organized based of global competition in such a way as to ensute uniform mapping errors. The cooperative interpolation among the multiple clusters associated with a mapping allows the network to perform a bidirectional many-to-many mapping, representing a particular from of defuzzification. Finally, an inference engine is constructed for the network to search for an optimal chain of rules or situation transitions under the constraint of transition feasibilities imposed by the learned mapping. Applications of the proposed network to skill acquisition are shown.

• Journal of KIISE
• /
• v.44 no.8
• /
• pp.774-781
• /
• 2017

Mention detection systems use nouns or noun phrases as a head and construct a chunk of text that defines any meaning, including a modifier. The term "mention detection" relates to the extraction of mentions in a document. In the mentions, a coreference resolution pertains to finding out if various mentions have the same meaning to each other. A pointer network is a model based on a recurrent neural network (RNN) encoder-decoder, and outputs a list of elements that correspond to input sequence. In this paper, we propose the use of mention detection using pointer networks. Our proposed model can solve the problem of overlapped mention detection, an issue that could not be solved by sequence labeling when applying the pointer network to the mention detection. As a result of this experiment, performance of the proposed mention detection model showed an F1 of 80.07%, a 7.65%p higher than rule-based mention detection; a co-reference resolution performance using this mention detection model showed a CoNLL F1 of 52.67% (mention boundary), and a CoNLL F1 of 60.11% (head boundary) that is high, 7.68%p, or 1.5%p more than coreference resolution using rule-based mention detection.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.5 no.12
• /
• pp.2355-2373
• /
• 2011

Vision-based vehicle detector systems are becoming increasingly important in ITS applications. Real-time operation, robustness, precision, accurate estimation of traffic parameters, and ease of setup are important features to be considered in developing such systems. Further, accurate vehicle detection is difficult in varied complex traffic environments. These environments include changes in weather as well as challenging traffic conditions, such as shadow effects and jams. To meet real-time requirements, the proposed system first applies a color background to extract moving objects, which are then tracked by considering their relative distances and directions. To achieve robustness and precision, the color background is regularly updated by the proposed algorithm to overcome luminance variations. This paper also proposes a scheme of feedback compensation to resolve background convergence errors, which occur when vehicles temporarily park on the roadside while the background image is being converged. Next, vehicle occlusion is resolved using the proposed prior split approach and through reasoning for rule-based tracking. This approach can automatically detect straight lanes. Following this step, trajectories are applied to derive traffic parameters; finally, to facilitate easy setup, we propose a means to automate the setting of the system parameters. Experimental results show that the system can operate well under various complex traffic conditions in real time.

• ETRI Journal
• /
• v.9 no.1
• /
• pp.84-96
• /
• 1987

The equality relation is very important in mechanical theorem proving procedures. A proposed inference rule called RHU-resolution is intended to extend the hyperparamodulation[23, 9] by introducing a bidirectional proof search that simultaneously employs a forward reasoning and a backward reasoning, and generalize it by incorporating beneflts of extended hyper steps with a preprocessing process, that includes a subsumption check in an equality graph and a high level planning. The forward reasoning in RHU-resolution may replace the role of the function substitution link.[9] That is, RHU-deduction without the function substitution link gets a proof. In order to control explosive generation of positive equalities by the forward reasoning, we haue put some restrictions on input clauses and k-pd links, and also have included a control strategy for a positive-positive linkage, like the set-of-support concept, A linking path between two end terms can be found by simple checking of linked unifiability using the concept of a linked unification. We tried to prevent redundant resolvents from generating by preprocessing using a subsumption check in the subsumption based eauality graph(SPD-Graph)so that the search space for possible RHU-resolution may be reduced.

• KIISE Transactions on Computing Practices
• /
• v.23 no.9
• /
• pp.542-549
• /
• 2017

Sequence-to-sequence models and similar pointer networks suffer from performance degradation when an input is composed of multiple sentences or when the length of the input sentence is long. To solve this problem, this paper proposes a hierarchical pointer network model that uses both the word level and sentence level information to encode input sequences composed of several sentences at the word level and sentence level. We propose a hierarchical pointer network based coreference resolution that performs a coreference resolution for all mentions. The experimental results show that the proposed model has a precision of 87.07%, recall of 65.39% and CoNLL F1 74.61%, which is an improvement of 21.83% compared to an existing rule-based model.

• Journal of KIISE
• /
• v.44 no.5
• /
• pp.496-502
• /
• 2017

Pointer Networks is a deep-learning model for the attention-mechanism outputting of a list of elements that corresponds to the input sequence and is based on a recurrent neural network (RNN). The coreference resolution for pronouns is the natural language processing (NLP) task that defines a single entity to find the antecedents that correspond to the pronouns in a document. In this paper, a pronoun coreference-resolution method that finds the relation between the antecedents and the pronouns using the Pointer Networks is proposed; furthermore, the input methods of the Pointer Networks-that is, the chaining order between the words in an entity-are proposed. From among the methods that are proposed in this paper, the chaining order Coref2 showed the best performance with an F1 of MUC 81.40 %. The method showed performances that are 31.00 % and 19.28 % better than the rule-based (50.40 %) and statistics-based (62.12 %) coreference resolution systems, respectively, for the Korean pronouns.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.30 no.4
• /
• pp.397-404
• /
• 2012

Jeju where concentrates on agriculture and tourism, conversion of outdoor culture into cultivation under structure happens actively for the purpose of increasing profit so continuous examination on house cultivation area is very important for this region. This paper is to suggest the effective image classification method using high resolution satellite image to detect the greenhouse. We carried out classification of greenhouse using the supervised classification and rule-based classification method about Formosat-2 images. Connecting result of two classification try to find accuracy improvement for greenhouse detection. Results about each classification method were calculated the accuracy by comparing with the result of visual detection. As a result, mahalanobis distance among the supervised methods was resulted in the highest detection. Also, it could be checked that detection accuracy was improved by tying with result of supervised method and result of rule-based classification. Therefore, it was expected that effective detection of greenhouse would be feasible if henceforward further study is performed in the process of connecting supervised classification and rule-based classification.

• The Transactions of the Korea Information Processing Society
• /
• v.3 no.7
• /
• pp.1792-1802
• /
• 1996

The analysis of prepositional phrases in English-to Korean machine translation has problem on the PP-attachment resolution, semantic analysis, and acquisition of information. This paper presents an analysis system for prepositional phrases, which solves the problem. The analysis system consists of the PP-attachment resolution hybrid system, semantic analysis system, and semantic feature generator that automatically generates input information. It provides objectiveness in analyzing prepositional phrases with the automatic generation of semantic features. The semantic analysis system enables to generate natural Korean expressions through selection semantic roles of prepositional phrases. The PP-attachment resolution hybrid system has the merit of the rule-based and neural network-based method.

• The KIPS Transactions:PartD
• /
• v.9D no.6
• /
• pp.991-998
• /
• 2002

With the advent of XML and database languages armed with the object-oriented concept and deductive logic, the problem of efficient query processing for them has become a major issue. We describe a way of processing semi-structured XML data through an implementation of a Deductive and Object-oriented Database (DOODB) language with the explanation of query processing. We have shown how to convert an XML data model to a DOODB data model. We have then presented an efficient query processing method based on Connection Graph Resolution. We also present a knowledge-based query processing method that uses the homomorphism of objects in the database and the associative rule of substitutions.