1 2538-4201 Research Center on Developing Advanced Technologies 182 Paper Proposing an intelligent and semantic-based system for Evaluating Text Summarizers Tabatabaei Raziyeh Feizi-Derakhshi Mohammad-Reza c Masoumi Saeid c University of Tabriz 1 9 2015 12 2 3 11 19 11 2013 07 04 2015 Nowadays summarizers and machine translators have attracted much attention to themselves, and many activities on making such tools have been done around the world. For Farsi like the other languages there have been efforts in this field. So evaluating such tools has a great importance. Human evaluations of machine summarization are extensive but expensive. Human evaluations can take months to finish and involve human labor that cannot be reused. In this paper, we propose a method of automatic machine summarization evaluation that is quick, inexpensive, and language-independent, that correlates highly with human evaluation, and that has little marginal cost per run. This method has the metrics of determining auto summaries’ quality, through comparing them to the summaries produced by Human (ideal summaries). These metrics measures overlapping of system summaries and human ones in number of units like n-tuples, words string and pairs of words. Certainly for semantic comparing of texts in case of review summaries, the appearance of words are not enough and using of WordNet seems to be necessary. In the proposed method words network is used with an appropriate idea and has improved evaluation results significantly. The proposed method is the first method for the Persian language. Performance measurement of the tool was done during a specified and standard procedure and the results indicate acceptable yield of it. We present this method as an automated understudy to skilled human judges which substitutes for them when there is need for quick or frequent evaluations.
45 Paper Contrast Relation Recognition in Persian discourse using supervised learning methods khodadadi habib e rahati quchani saeed f estaji azam g e islamic azad university f islamic azad university g ferdowsi university 1 9 2015 12 2 13 22 22 05 2013 13 02 2015 Discourse is a part of language that intend is used to communicate. A discourse relation recognition system can identify one or more relation between the textual units in a discourse. Like other languages, Contrast relation is a one of the available relations in Persian discourse. Contrast relation recognition in discourse is useful for generation and perception of discourse, paraphrasing and summarization systems and et al. This relation in one discourse is often detected by discourse marker such as “اما” and “ولی”, But in some situation these markers are removed and relation recognition is difficult. For this reason we have proposed to use of some feature for relation recognition. These features are: tense of verbs, word pairs, and et al. In this paper, a Corpus of Research Center of Intelligent Signal Processing has been used to collect 5000 instances of contrast and 5000 other relations then created feature vector for each instance. We used three supervised methods for classification: SVM, KNN ,Parzen Window and combine of this classifiers. Finally, the best result achieved by combine classifier that accuracy is 87.13. 185 Paper Two Featuer Transformation Methods Based on Genetic Algorithm for Reducing Support Vector Machine Classification Error hoseinkhani fatemeh h nasersharif babak i h Qazvin Islamic Azad University i K.N Toosi University of Technology 1 9 2015 12 2 23 39 24 11 2013 08 12 2014 Discriminative methods are used for increasing pattern recognition and classification accuracy. These methods can be used as discriminant transformations applied to features or they can be used as discriminative learning algorithms for the classifiers. Usually, discriminative transformations criteria are different from the criteria of  discriminant classifiers training or  their error. In this paper, for relating feature transformation criterion to classification rate, we obtain a feature transformation method using genetic algorithm where we choose fitness function as Support Vectomr Machine(SVM) classification error rate. In addition, we obtain a feature transformation method using multi-objective genetic algorithm in order to consider both between class discrimination (According to feature transformation criterion) and support vector machine classification error rate simultaneously. Experimental results on UCI dataset indicate that using both classification error and between class discrimination in feature transformation improve discriminative feature transformations performance in increasing SVM classification accuracy. Additionally, the use of feature transformation with classification error criterion increases SVM classification more than other conventional feature transformation and proposed two-objective methods. 198 Paper EEG investigation of the effective brain networks for recognizing musical emotions Hasanzadeh Fatemeh j Shahabi Hossein k Moghimi Sahar l Moghimi Ali m j Ferdowsi University of Mashhad k Ferdowsi University of Mashhad l Ferdowsi University of Mashhad m Ferdowsi University of Mashhad 1 9 2015 12 2 41 54 04 01 2014 07 11 2014 In the current research brain effective networks related to happy and sad emotions are studied during listening to music. Connectivity patterns among different EEG channels were extracted using multivariate autoregressive modeling and partial directed coherence while participants listened to musical excerpts. Both classical and Iranian musical selections were used as stimulus. Participants’ self-reported emotional values were used for classification of excerpts. The connectivity matrices varied from happy to sad musical selections. Moreover, the parameters extracted from different regions correlated with subjective assessments of the emotional content. Self-reported valance had a positive correlation with the inflow of frontal channels while listening to happy excerpts. This correlation was negative for sad pieces. The obtained results demonstrate that the connectivity indices among different regions can be used for differentiating happy and sad emotions. 190 Paper Extracting parallel corpora from web comparable documents to improve the quality of an English-Farsi translation system rahimi zeinab n samani mohammad hossein o khadivi shahram p n RCISP o RCISP p amirkabir univercity of technology, department of computer engineerin 1 9 2015 12 2 55 72 12 12 2013 25 08 2014 Data used for training statistical machine translation method are usually prepared from three resources: parallel, non-parallel and comparable text corpora. Parallel corpora are an ideal resource for translation but due to lack of these kinds of texts, non-parallel and comparable corpora are used either for parallel text extraction. Most of existing methods for exploiting comparable corpora look for parallel data at the sentence level. However, we believe that very non-parallel corpora have none or few good sentence pairs most of their parallel data exists at the sub-sentential level. The base system is Manteanu 2006 fragment extraction system implemented in C# and the proposed system is implemented based on extracting fragment blocks from input related sentences using score calculated from special features such as fragment length, LLR score, relevance path specification in the block and translation coverage percent. Evaluations indicates that proposed method outperforms the base system and the improved base system. 36 Paper Combination of event related potentials and Peripheral signals in order to improve the accuracy of the Lie detection Systems Ghodousi mahrad Nasrabadi Ali moti Torabi Shahla Mohammadian Amin Mehrnam AmirHossein Rcdat, Shahed University Shahed University Rcdat Rcdat Shahed University 1 9 2015 12 2 73 86 01 05 2013 24 06 2015 Since it was being predicted that combination of psychophysiological and ERP signals, during the detection of a guilty person's knowledge can increase the performance of integrative lie detection system toward using the separate procedures Using the knowledge of both aspects, in this study it has been tried to determine the proper Inter-Stimulus Intervals (ISI) together with suitable sequence of stimulations in order to simultaneous recording of P300 component of brain Event Related Potentials and peripheral signals. Also a proper mock crime scenario has been designed it has the capability of exciting the cognitive aspect of mock crime and also was capable of provoking the subject’s concerns, based on telling lie about the crime. At the next stage, after recording data from 32 participants, features from their ERP and SCR (as one of the most important peripheral signals) signals have been extracted. Then, an LDA classifier was applied on selected features which were selected by Genetic algorithm and these accuracies: 76.67%, 73.33% & 80% have been obtained for EEG, SCR and Combined data respectively. The resulted accuracies at the first show the proper quality of scenario and protocol, in synchronous stimulation and recording of both signal categories, also the improvements which have been resulted by integrative data in compare with separate ones are observable. 203 Paper Extractive summarization based on cognitive aspects of human mind for narrative text 1 9 2015 12 2 87 96 11 01 2014 21 04 2015 This study explains a summarization system based on a cognitive model theory. This theory is about comprehension and is used to explain comprehending narrative texts. Majority of previous methods have been used statistical approaches for summarization, and this method is different as it tries to build a system based on a cognitive theory and not statistical methods. Main principle of situational model is that as humans read a text, they will make a mental image based on temporality, causality, intentionality, protagonists, and place. Proposed system extracts five features for each sentence and identifies the most important sentences based on five features. The results obtained from this method were satisfactory. 54 Paper Speech Enhancement Using MMSE Estimator Based on Mixture of laplacian 1 9 2015 12 2 97 107 03 06 2013 04 08 2015 In this paper an estimator of speech spectrum for speech enhancement based on Laplacian Mixture Model has been proposed. We present an analytical solution for estimating the complex DFT coefficients with the MMSE estimator when the clean speech DFT coefficients are mixture of Laplacians distributed. The distribution of the DFT coefficients of noise are assumed zero-mean Gaussian.The drived MMSE estimator is non-linear and it was shown that this estimator performs better than estimators which are based on Gaussian and Laplacin model