Informace o publikaci
Extraction of hidden topics in urban context based on the Internet publications analysis
| Autoři | |
|---|---|
| Rok publikování | 2022 |
| Druh | Článek ve sborníku |
| Konference | Procedia Computer Science |
| Fakulta / Pracoviště MU | |
| Citace | |
| www | https://doi.org/10.1016/j.procs.2022.10.204 |
| Doi | http://dx.doi.org/10.1016/j.procs.2022.10.204 |
| Klíčová slova | named entities recognition; Natural language processing; texts clustering; topic modeling; urban environment; urban environment object |
| Popis | The problem considered in the article is the systematic lack of data on the objects of the urban environment for management and decision-making. This problem is particularly acute in the lack of data on points of attraction, informal and thematic places of interest. At the same time, this kind of information is necessary for the qualitative development of the urban environment. This article discusses an approach to creating new information resources based on the analysis of publications and messages of citizens, which can be used to effectively manage the development of the city and improve the quality of the urban environment. For example, to create new centers of attraction for citizens and tourists and effective landscaping. Currently, there is a public demand for the semantic content of the urban environment, considering historical and cultural associations, informal symbols. Traditionally, this request is met through surveys of the population in urban improvement projects. This article presents an approach to supplementing such surveys with information from Internet social networks processed by natural language analysis methods to extract hidden topics and thematic objects of the urban environment. The approach is demonstrated based on the example of the city of St. Petersburg in the Russian Federation. |