We propose an in-depth study of lending behaviors in Kiva using a mix of quantitative and large-scale data mining techniques. Kiva is a non-profit organization that offers an online platform to connect lenders with borrowers. Their site, kiva.org, allows citizens to microlend small amounts of money to entrepreneurs (borrowers) from different countries. The borrowers are always affiliated with a Field Partner (FP) which can be a microfinance institution (MFI) or other type of local organization that has partnered with Kiva. Field partners give loans to selected businesses based on their local knowledge regarding the country, the business sector including agriculture, health or manufacture among others, and the borrower.Our objective is to understand the relationship between lending activity and various features offered by the online platform. Specifically, we focus on two research questions: (i) the role that MFI ratings play in driving lending activity and (ii) the role that various loan features have in the lending behavior. The first question analyzes whether there exists a relationship between the MFI ratings - that lenders can explore online - and their lending volumes. The second research question attempts to understand if certain loan features - available online at Kiva - such as the type of small business, the gender of the borrower, or the loan's country information might affect the way lenders lend.
翻译:我们提议对Kiva的贷款行为进行深入研究,利用定量和大规模数据开采技术的组合进行。Kiva是一个非营利组织,提供在线平台,将贷款人与借款人连接起来。Kiva.org网站允许公民向不同国家的企业家(借款人)提供小额资金。借款人始终与外地伙伴(FP)有联系,该伙伴可以是小额供资机构(MIF)或与Kiva合作的其他类型的当地组织。外地伙伴根据他们对国家、商业部门(包括农业、保健或制造业等)和借款人的当地知识,向选定的企业提供贷款。我们的目标是了解贷款活动与在线平台提供的各种特征之间的关系。具体地说,我们侧重于两个研究问题:(一) Mfi公司评级在推动贷款活动方面所起的作用,以及(二)各种贷款特征在贷款行为中所起的作用。第一个问题分析MFI的评级――贷款人可以在线探索――及其贷款量之间的关系。第二个研究问题试图了解,如果贷款人的某些贷款特点――在Kiva这类类型的借款人可能影响到贷款国的性别信息。