Normal view MARC view ISBD view

Using Probabilistic Logic to Model the Impact of Entrepreneurs’ Social Network Reach on the Survival Rate of Startup Businesses / Lauc, Davor ; Grgić, Siniša ; Boras, Damir.

By: Lauc, Davor.
Contributor(s): Grgić, Siniša [aut] | Boras, Damir [aut].
Material type: ArticleArticleDescription: 24-29 str.Other title: Using Probabilistic Logic to Model the Impact of Entrepreneurs’ Social Network Reach on the Survival Rate of Startup Businesses [Naslov na engleskom:].Subject(s): 5.01 | SNA; Business Startups; Probabilistic Logic | SNA; Business Startups; Probabilistic Logic In: 2nd International Conference on Economics and Social Sciences str. 24-29Summary: It is generally widely accepted within business circles that the founders’ ability to network is one of the key factors that underpins the success of a startup. However, this belief is not well researched and is not founded on empirical real- life data. Our research is based on a large-scale longitudinal study of more than 5, 000 Croatian start-up businesses. Data on these businesses’ results and the social networks of their founders were collected from public sources and interviews with entrepreneurs over a period of 10 years. Standard methods of social network analysis were applied to evaluate founders’ egocentric social networks and their position in the larger business social network. The key indicators of the presence of an ego-network that were applied within the research included structural measures (degree/density, efficiency) and compositional measures (number of entrepreneurs in the network).  Key indicators of the founders’ position in the wider business social network included degree, closeness, betweenness and eigenvector centrality. Analysis demonstrated that almost all indicators showed a measurable and statistically significant connection between entrepreneurs’ social network reach and the business results of their companies, especially in terms of survival rate. The dataset were further used to build a predictive model of the probability of start-up business survival based on its founder’s social network reach. The model was developed by machine learning techniques that were based on the Markov Logic Networks framework. Preliminary results indicated that the model demonstrates significant predictive power and accuracy.
Tags from this library: No tags from this library for this title. Log in to add tags.
No physical items for this record

It is generally widely accepted within business circles that the founders’ ability to network is one of the key factors that underpins the success of a startup. However, this belief is not well researched and is not founded on empirical real- life data. Our research is based on a large-scale longitudinal study of more than 5, 000 Croatian start-up businesses. Data on these businesses’ results and the social networks of their founders were collected from public sources and interviews with entrepreneurs over a period of 10 years. Standard methods of social network analysis were applied to evaluate founders’ egocentric social networks and their position in the larger business social network. The key indicators of the presence of an ego-network that were applied within the research included structural measures (degree/density, efficiency) and compositional measures (number of entrepreneurs in the network).  Key indicators of the founders’ position in the wider business social network included degree, closeness, betweenness and eigenvector centrality. Analysis demonstrated that almost all indicators showed a measurable and statistically significant connection between entrepreneurs’ social network reach and the business results of their companies, especially in terms of survival rate. The dataset were further used to build a predictive model of the probability of start-up business survival based on its founder’s social network reach. The model was developed by machine learning techniques that were based on the Markov Logic Networks framework. Preliminary results indicated that the model demonstrates significant predictive power and accuracy.

Projekt MZOS projekt

ENG

There are no comments for this item.

Log in to your account to post a comment.

Powered by Koha

//