Text Mining of Donor's conversation with Solicitors

Track: Data Analytics

Session Number: 1109
Date: Thu, Aug 9th, 2018
Time: 3:00 PM - 3:30 PM

Description:

Solicitors pursue alumni all the time into giving back to the school. Solicitors contact alumni via all the means and note down every conversation they have with them, which are summarized conversations with personal notes added. The objective of this paper is to analyze the text to uncover insights and in the end do predictive modelling. Using text mining nodes in SAS Enterprise Miner 14.1, we will perform text analytics on a database obtained from a university foundation that contained 38,000 observations.
In text analytics, text clustering was conducted and meaningful clusters were obtained and utilized for donor segmentation. We also used text topic node. The rule builder node was used to find key words that were associated with a donation.
Predictive modelling was conducted on text data and text/ numerical combinations and various models were compared. The numeric variables were three internal ratings, gender, degree, school, marital status and state. We would be seeing if textual data when combined with numeric data outperforms numeric data alone or textual data alone.
This paper is going to benefit any fundraising organization and widen the scope of their methods and the way they reach out to constituents.
Session Type: PD or DAS Case Study (30 minutes)

Primary Competency: DA:Competency 1: Cross Industry Standard Process for Data Mining (CRISP-DM)
Secondary Competency: DA:Competency 8: Analytics and Campaign
Tertiary Competency: DA:Competency 3: Data Manipulation Skills
Intended Audience Level: Level I, Level II
Recommended Prerequisites: none
Learning Objective #1: Attendees will be able to understand what to with the conversations recorded between donors and solicitors.
Learning Objective #2: Attendees will be able to learn some of the basic predictive modeling techniques.
Shop Size: All Shop Sizes
Session Type: PD or DAS Case Study (30 minutes)

Primary Competency: DA:Competency 1: Cross Industry Standard Process for Data Mining (CRISP-DM)
Secondary Competency: DA:Competency 8: Analytics and Campaign
Tertiary Competency: DA:Competency 3: Data Manipulation Skills
Intended Audience Level: Level I, Level II
Recommended Prerequisites: none
Learning Objective #1: Attendees will be able to understand what to with the conversations recorded between donors and solicitors.
Learning Objective #2: Attendees will be able to learn some of the basic predictive modeling techniques.
Shop Size: All Shop Sizes