Building Trusted Startup Teams from LinkedIn Attributes: A Higher Order Probabilistic Analysis

Document Type

Conference Proceeding

Source of Publication

Proceedings - International Conference on Tools with Artificial Intelligence, ICTAI

Publication Date

11-1-2020

Abstract

© 2020 IEEE. Startups arguably contribute to the current business landscape by developing innovative products and services. The discovery of business partners and employees with a specific background which can be verified stands out repeatedly as a prime obstacle. LinkedIn is a popular platform where professional milestones, endorsements, recommendations, and skills are posted. A graph search algorithm with a BFS and a DFS strategy for seeking trusted candidates in LinkedIn is proposed. Both strategies rely on a metric for assessing the trustworthiness of an account according to LinkedIn attributes. Also, a stochastic vertex selection mechanism reminiscent of preferential attachment guides search. Both strategies were verified against a large segment of the vivid startup ecosystem of Patras, Hellas. A higher order probabilistic analysis suggests that BFS is more suitable. Findings also imply that emphasis should be given to local networking events, peer interaction, and to tasks allowing verifiable credit for the respective work.

ISBN

9781728192284

ISSN

1082-3409

Publisher

IEEE

Volume

2020-November

First Page

867

Last Page

874

Disciplines

Computer Sciences

Keywords

graph mining, higher order statistics, linked data, LinkedIn API, multilayer graphs, probabilistic analysis, trust

Scopus ID

85098790570

Indexed in Scopus

yes

Open Access

no

Share

COinS