Loading...
Please wait, while we are loading the content...
Similar Documents
Donor Sentiment and Characteristic Analysis Using SAS® Enterprise Miner™ and SAS® Sentiment Analysis Studio
| Content Provider | Semantic Scholar |
|---|---|
| Author | Kakarla, Ramcharan Chakraborty, Goutam |
| Copyright Year | 2015 |
| Abstract | It has always been a million-dollar question, “What inhibits a donor to donate?” Many successful universities have deep roots in annual giving. We know sentiment is a key factor in drawing attention to engage donors. This paper is a summary of findings about donor behaviors using textual analysis combined with the power of predictive modeling. In addition to identifying the characteristics of general donors, the paper focuses on identifying the characteristics of a first-time donor. It distinguishes the features of the first-time donor from the general donor pattern. A data set containing 247,000 records was obtained from a University Foundation alumni database, Facebook, and Twitter. Solicitation content such as email subject lines sent to the prospect base was considered. Time-dependent data and time-independent data were categorized to make unbiased predictions about the first-time donor. The predictive models use inputs such as age, educational records, scholarships, events, student memberships, and solicitation methods. Models such as decision trees, Dmine regression, and neural networks were built to predict the prospects. SAS® Sentiment Analysis Studio and SAS® Enterprise Miner™ were used to analyze the sentiment. |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | https://business.okstate.edu/site-files/docs/analytics/3347-2015.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
