<p>The present study analyzes the naming of law firms in Russia, with a particular focus on the geographical aspects of this phenomenon. The main objectives of this research are to identify law firm naming patterns, to categorise them and to test whether there are any spatial regularities in the distribution of law firms according to the identified patterns. To conduct the analysis, data regarding law firms is extracted from the open dumps of the state registry of small and medium-sized enterprises. The names of the firms are transformed into numerical vectors, subsequently clustered with the KMeans method, and then manually grouped to identify naming patterns and strategies. Global and local Moran’s <i>I</i> statistics are calculated to test for spatial autocorrelation in the proportion of a given pattern in the regions. As a result, 22 naming patterns are identified, grouped into four naming strategies. Apart from no-strategy (40% of law firms), the most popular strategy is belonging (to the legal services market—30%), followed by uniqueness (of service or location—20%) and market (words and phrases with positive or promotional meaning—10%). The research findings indicate that the spatial distribution of law firm naming patterns is generally not significantly different from random, with two exceptions: geographical branding in Siberia and an unusual heighborhood of regions with high and low proportions of “Expert” and “Personal brand” patterns in the European South.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Spatial distribution of naming patterns in Russian law firms through text embeddings

  • Pavel Syomin

摘要

The present study analyzes the naming of law firms in Russia, with a particular focus on the geographical aspects of this phenomenon. The main objectives of this research are to identify law firm naming patterns, to categorise them and to test whether there are any spatial regularities in the distribution of law firms according to the identified patterns. To conduct the analysis, data regarding law firms is extracted from the open dumps of the state registry of small and medium-sized enterprises. The names of the firms are transformed into numerical vectors, subsequently clustered with the KMeans method, and then manually grouped to identify naming patterns and strategies. Global and local Moran’s I statistics are calculated to test for spatial autocorrelation in the proportion of a given pattern in the regions. As a result, 22 naming patterns are identified, grouped into four naming strategies. Apart from no-strategy (40% of law firms), the most popular strategy is belonging (to the legal services market—30%), followed by uniqueness (of service or location—20%) and market (words and phrases with positive or promotional meaning—10%). The research findings indicate that the spatial distribution of law firm naming patterns is generally not significantly different from random, with two exceptions: geographical branding in Siberia and an unusual heighborhood of regions with high and low proportions of “Expert” and “Personal brand” patterns in the European South.