Применение рекомендательных систем на базе социальных сетей для изучения иностранных языков
Building of adequate dynamical models of microblogging social networks is a topical task that is of interest from both theoretical and practical aspects. Experimental and theoretical results of studies related to choice of the adequate model are presented. The choice was made between two models: a nonlinear dynamical system and a nonlinear random dynamical system. By results of the fractal analysis of observable network time series and defining their probability density function it was established that the nonlinear random dynamical system was more adequate than the nonlinear dynamical system. The character of the observable time series was also explored. The possibility that microblogging social networks can be analyzed by means of Tsallis entropy and self-organized criticality is examined.
A general model of socio-semantic network is presented in terms of state-transitions systems. We provide some examples and indicate research directions, which seem to us the most important from the application point of view.
In this paper we focus on the problem of user prediction in visual product recommender systems based on the given set of photos of products purchased by the user previously. We studied neural aggregation methods for image features extracted by the deep neural networks. We propose the novel two-stage algorithm. At first, the image features are learned by fine-tuning the convolutional neural network. At the second stage, we sequentially combine the known learnable pooling techniques (neural aggregation network and context gating) in order to compute a single descriptor for particular user as a weighted average of image features. It is experimentally shown for the Amazon product dataset that F1-measure for our approach is more than 20% higher when compared to conventional averaging of the feature vector.
The article presents the development of the ontology for a multi-agent subsystem analysing user posts in social networks in order to identify security threats to society. The testing of multi-agent subsystem using the developed ontology is described.
Comparative analysis of network and real-life identity explores two hypotheses: 1) aspects of identity, its different parameters may have diverse profiles for the network and the reality conditions; 2) they may also indicate gender and age differences. The study is held on the sample of 42 participants, aged from 15 to 25, who were interviewed. Gender and age differences were found referring to the social identity in the network and the reality, and for superficial identity in the network condition, as well as differences for individual and relational identity in the network and the reality conditions. Variability of the factor structure was found for the network condition in comparison with the aspects of identity in reality.