The Extension of Persian Search Engine Based on Improving
Ontology Learning Process
Sima
Darvishi
Department of Computer Engineering, University of Guilan
author
Asadollah
Shahbahrami
Department of Computer Engineering, University of Guilan
author
Manouchehr
Nahvi
Department of Electrical Engineering, University of Guilan
author
text
article
2014
per
A search engine is a kind of tool which fulfills users' information needs. Users try to obtain thedesirable objectives adopting different queries in search engines. Some queries are carried out throughinserting effective keywords. If a search engine fairly understands the relation among words, the userscan certainly extract better results. However, understanding these relations and expressing the user'sintention relatively refer to the structure of queries in any language. It is not easy to perform searchsin a Persian search engine according to the syntactic, pronunciation and spelling rules. In order toenhance the accuracy of the Persian search engines, the concept of ontology can be used to describewords and understanding their concepts. This study presented a model for improving the extraction ofsemantic relations from Persian language in terms of the performance of search engine and ontology.In this model, a standard Persian text collection named Bijankhan corpus is used. Persian languageand syntactic patterns obtained from experiments on some texts of this corpus indicated that theaccuracy of the proposed model is about 87%.
Journal of Soft Computing and Information Technology
Babol Noshirvani University of Technology
2383-1006
3
v.
3
no.
2014
12
29
https://jscit.nit.ac.ir/article_101599_1af92e43ac807c598529c1df0208f520.pdf
Banking Facility Placement in a City Using an Evolutionary Approach
Fatemeh
Karimi
Departemen of Computer Science, University of Tabriz, Tabriz, Iran
author
Shahriar
Lotfi
Departemen of Computer Science, University of Tabriz, Tabriz, Iran
author
text
article
2014
per
Facility location problems are classical optimization problems that have numerous applications,especially in the service industries. Banks as part of the firms that are in contact with people everyday, have especial sensitivity in choosing appropriate location to maximize their market share andincrease customer satisfaction by providing quick access. Also by considering competitive pressurebetween various banks to attract potential customers, optimal placement of banking facility hasespecial importance for managers. Finding banking facilities location is determined by using theMaximal Covering Location Problem model with the aim of covering maximal commercial potentialwith considering some criteria. Current research, which has been done based on the Tejarat bankrequirements for a practical project, modifies geographic distribution of banking facilities using anevolutionary algorithm with new features. The proposed algorithm, called TPCEA, whit a newencoding method and the use of different operators tries to achieve an optimal configuration forbanking facilities. Results show the power of algorithm to achieve optimal solutions, appropriatescalability to increase the size of the problem, acceptable convergence and its high stability indifferent conditions.
Journal of Soft Computing and Information Technology
Babol Noshirvani University of Technology
2383-1006
3
v.
3
no.
2014
30
47
https://jscit.nit.ac.ir/article_99005_61d2c2366c4baad6f69018bb1350220e.pdf
A New Fuzzy-Genetic Filter for Digital Image Noise Reduction
Somayeh
Aliakbari
Computer Engineering Department, Yazd University, Yazd, Iran.
author
Mohammad
Ghasemzadeh
Computer Engineering Department, Yazd University, Yazd, Iran.
author
Ali Mohammad
Latif
Computer Engineering Department, Yazd University, Yazd, Iran.
author
text
article
2014
per
In this paper we present a novel approach for digital image noise reduction. Fuzzy Systems and Genetic Algorithms have been already used in image noise reduction, innovation of the proposed method relies on transmission of image pixels into a noise-free environment, also, determination of set fuzzy rules and optimal parameters of this system by utilizing genetic algorithms. Necessity and importance of performing this research work leans on the great demand for new methods that remove noise while preserve details of the original image. In the proposed method, We find median values of eight neighborhood cells of the given image pixels to determine input parameters of the fuzzy system. We also utilize a genetic algorithm to find the set rules and parameters of the membership function of our system automatically and intelligently. Chromosomes of this algorithm are fuzzy rules and parameters of the membership functions. We implemented the proposed filter tested it over standard images. Results reveal the ability of this method in removing noise while preserving details of the image. Logical arguments and experimental results show that the mainadvantage of this filter is its ability to remove high-density noise.
Journal of Soft Computing and Information Technology
Babol Noshirvani University of Technology
2383-1006
3
v.
3
no.
2014
48
58
https://jscit.nit.ac.ir/article_96202_8126e8a161eebd10d54e14195d2212e1.pdf