Data Mining of a Small Text-based Dataset

Project Description

Data mining with RapidMiner

Doctoral candidate requesting a Rapid Miner professional experienced with text data mining. Must be certified and will need to provide evidence of certification.

Looking for help with trend/topic detection in a ~2000 record dataset. 1 column/field. As many as 400-500 words in a record. See attached data sample.

The data is extracted from a Help Desk database. I'm interested in data mining, trend/topic detection for column F (Description).

Looking for someone to do the mining tasks. I can do data cleansing. I'm looking for a partner that I can discuss the project with. I'll tell you what I'm trying to do, you tell me how you need the data cleaned and prepped, then I'll tell you what (I think) I need from mining round 1. You provide the results and details of the methodology, I'll review and request a second round. We'll do this a few times. My ultimate goal is something like a ranked list of topics reported to the help desk.

As this is for a doctoral dissertation, I'll need clear details on exactly what steps you executed.

Data is .csv or Excel, or anything else you need. Attached example is in Excel.

Deliverables are the results of the initial analyses and some secondary analyses. Final deliverable is a list of trends across the records in the dataset, with frequency (etc) information. Excel or Word or whatever format you like. Will need to discuss analyses and results with the miner.

When providing proposal, please consider this is a student project, and not a commercial project. Please submit proposal accordingly. 

I am a doctoral candidate and paying for this out of pocket. It's not a huge pay day but it's not much work either! I can provide a research assistant credit if you'd like. (Dates in the posting are approximate).