Automated data collection for credit score calculation based on financial transactions and social media
Automated data collection for credit score calculation based on financial transactions and social media
复制标题
基于金融交易和社交媒体的信用评分计算的自动数据收集
DOI：
10.1109/etiict.2017.7977024
复制
发表时间：
2017
期刊：
影响因子：
--
通讯作者：
S. Sontakke
中科院分区：
文献类型：
--
作者：
J. Lohokare;Reshul Dani;S. Sontakke
一个人的财务信誉是批准贷款或允许信贷交易的主要因素。如今，这种可信度基于一个人的“信用评分”，该评分是根据该人过去的债务履行情况计算得出的。本文提供了一种独特的替代解决方案来收集这些数据。利用当今几乎每个人都拥有智能手机的事实，可以有一个智能手机应用程序来收集所有此类数据并将其提交给官方机构。金融交易并不是可以确定一个人可信度的唯一参数。本文建议访问社交媒体数据以深入了解一个人的一般社会地位。如今，所有交易均由银行和其他机构通过短信传送给用户。因此，访问 SMS 将导致获取与所有此类交易相关的数据。拟议的解决方案将具有智能手机应用程序，该应用程序将通过短信捕获银行交易数据和与在线购买相关的数据。使用人工神经网络将能够根据收集的各种数据参数计算最终的可信度得分。本文的主要贡献是它提供了一个自动收集计算信用评分所需的所有数据的系统。这种方法的独特之处在于社交媒体数据的参与。这种方法比现有的解决方案更好，因为它将收集交易数据以外的数据，从而能够计算更有效的信用评分。
Financial credibility of a person is a cardinal factor in the approval of loans or permitting credit transactions. Today, this credibility is based on the ‘credit score’ of the person which is calculated from the person's past performance on debt obligations. This paper provides a unique alternative solution to collect this data. Harnessing the fact that almost everyone has smartphones today, there can be a smartphone application that collects all such data and submits it to the official body. Financial transactions are not the only parameters that can determine credibility of a person. This paper proposes accessing social media data to get insights into general social status of a person. Today, all transactions are conveyed by banks and other institutes to the users via SMS. Hence, having access to SMS will result to getting data related to all such transactions. The proposed solution will have smartphone application that will capture bank transaction data and data related to online purchases through SMS. Use of artificial neural networks will enable calculating the final credibility score based on the various data parameters collected. The primary contribution of this paper is that it provides a system to automatically collect all data required to calculate the credit score. What is unique in this approach is the involvement of data from social media. This approach is better than the existing solutions as it will collect data other than just transactional data thus enabling calculation of a more effective credit score.