IRANIAN COMMUNICATION AND INFORMATION TECHNOLOGY
Year:2020 | Volume:11 | Issue:41-42|Papers Count:6

In the era of the Internet, recognition of adult images is important to children's physical and mental protection. It is a challenge to recognize adult images with changes in the illumination and skin color. In this paper, we proposed a new method for solving illumination normalization with skin color classification in the diagnosis of the adult image. In this paper, the deep fuzzy neural network method is utilized to improve the illumination normalization of adult images, which has improved the recognization of adult images is utilized. Using Xception to dividing the images and reduce the illumination variations in each part separately, which makes it possible to reduce the illumination variation in the whole image without losing details. In addition, the advanced color combination algorithm based on Gaussian-KNN algorithm is used for skin color classification, a non-parametric method is used for classifications and regressions. Finally, the SVM algorithm is utilized for image classification. In this paper, 33, 000 different types of images are collected from the Internet. The results show that the proposed method of 1/3 has improved the accuracy of the recognization.

Today, the emergence and expansion of technologies that provide the widest possible connection have brought about significant changes in the private life and professional life of individuals. Correct implementation of information technology is the source of economic and cultural development and the promotion of quality of life through the exchange of information and the provision of public and private services. The purpose of this research is to present the model of information technology acceptance in Iranian ICT research centers. Employed experts worked on ICT projects in this research Statistical population. It was provided in one of the university searching centers. And so a convenience and purposeful Nonprobability sampling does (30 person). This paper examines the factors and parameters affecting the acceptance of information technology in ICT projects of the university research centers in the field of information technology. To collect the required data and information, a questionnaire was used & to analyze the data and information obtained from the questionnaires using the Spss 22 and Smart pls3 software. According to the calculations, the factors affecting the acceptance of information technology in university research centers can be divided into four categories: IT related factors, organizational factors, factors related to executive director and individual factors that are related to management (0. 497), IT (0. 460) and Individual factors (0. 457) have an impact on the individual acceptance of information technology respectively and organizational factors (0. 469) on the adoption of an IT organization.

Predicting stock prices by data analysts have created a great business opportunity for a wide range of investors in the stock markets. But the fact is difficulte, because there are many affective economic factors in the stock markets that they are too dynamic and complex. In this paper, two models are designed and implemented to identify the complex relationship between 10 economic factors on the stock prices of companies operating in the Tehran stock market. First, a Mamdani Fuzzy Inference System (MFIS) is designed that the fuzzy rules set of its inference engine is found by the Particle Swarm Optimization Algorithm (PSO). Then a Deep Learning model consisting of 26 neurons is designed wiht 5 hidden layers. The designed models are implemented to predict the stock prices of nine companies operating on the Tehran Stock Exchange. The experimental results show that the designed deep learning model can obtain better results than the hybridization of MFIS-PSO, the neural network and SVM, although the interperative ability of the obtained patterns, more consistent behavior with much less variance, as well as higher convergence speed than other models can be mentioned as significant competitive advantages of the MFIS-PSO model.

News reports in social media are presented with large volumes of different kinds of documents. The presented topics in these documents focus on different communities and person stances and opinions. Knowing the relationships among persons in the documents can help the readers to obtain a basic knowledge about the subject and the purpose of various documents. In the present paper, we introduce a method for detecting communities that includes the persons with the same stances and ideas. To do this, the persons referenced in different documents are clustered into communities that have related positions and stances. In the presented method. Community-based personalities are identified based on a friendship network as a base method. Then by using a genetic algorithm, the way that these communities are identified is improved. The criterion in the tests is rand index of detection of these communities. The experiments are designed based on real databases that published in Google News on a particular topic. The results indicate the efficiency and desirability of the proposed method.

Recommender systems are the systems that help users find and select their target items. Most of the available events for recommender systems are focused on recommending the most relevant items to the users and do not include any context information such as time, location. This paper is presented by the use of geographically tagged photo information which is highly accurate. The distinction point between this thesis and other similar articles is that this paper includes more context (weather conditions, users’ mental status, traffic level, etc. ) than similar articles which include only time and location as context. This has brought the users close to each other in a cluster and has led to an increase in the accuracy. The proposed method merges the Colonial Competitive Algorithm and fuzzy clustering for a better and stronger processing against using merely the classic clustering and this has increased the accuracy of the recommendations. Flickr dataset is used to evaluate the presented method. Results of the evaluation indicate that the proposed method can provide location recommendations proportionate to the users’ preferences and their current visiting location.

Polar diagram is a generalization of Voronoi diagram in which the angle is used as the metric. This Problem has many applications in visibility, image Processing, telecommunication, antenna, and Path Planning Problems. In recent years two kinds of Polar diagram have been proposed and appropriate algorithm have been Presented for some types of sites. Also, some algorithms has presented for kinetic data and dynamic states. In this Paper, it is assumed that the Pole is moving and an algorithm is presented that updates near Pole Polar diagram of sites with moving pole efficiently and in a sub linear time. In this approach, the Preprocessing time is 〖 O(n^4 log〗 _2⁡ 〖 n)〗 and updating time for diagram with each successive movement is 〖 O(log〗 _2⁡ 〖 n +k)〗 that k is the number of sites in region T which its site’ s regions may be changed.