11-20-2012 09:29 PM
I have a dataset with the following information: AnnouncementDate, company, Description. Announcement date is the date when company announced changed of their CEO. In my sample I have some companies announcing the resignation of their CEO and some companies announcing appointment of their CEO. The "Description" provides information on what announcement was made by the company. I want to create two columns: ResignCEOCo takes value 1 if the Company announced resignation of CEO and AppointCEOCo takes value 1 if company announced appointment of CEO. Is there any SAS code that would read the description and create the dummy variables? If yes, can someone please share with me the codes?
My current dataset looks as follows:
|Park City Group Inc.||15628||Park City Group Names New Acting Chief Financial Officer|
|Petroleum Geo Services ASA||15628||Petroleum Geo-Services Changes Chief Financial Officer|
|Richtree Inc.||16006||Richtree Announces Resignation of Colin West as Vice President and Chief Financial Officer|
|Identive Group, Inc.||15658||SCM Microsystems Inc. elected Andrew Warner as Vice President, Finance and Chief Financial Officer|
|Euronet Worldwide Inc.||15641||Euronet Worldwide Introduces Chief Financial Officer and Office of Controller|
|MediSolution Ltd.||15641||MediSolution Announces New Chief Financial Officer|
|First Aviation Services Inc.||16005||First Aviation Services Inc. Names New Chief Financial Officer|
|Molex Inc.||15993||Molex Announces First Quarter 2003 Financial Results; Announces Second Quarter 2003 Earnings Guidance; Names Diane Bullock As Chief Financial Officer|
|Firestone Ventures Inc.||16001||Firestone Ventures Announces Appointment of Chief Financial Officer|
I want my new dataset to look as follows:
|Park City Group Inc.||15628||0||1|
|Petroleum Geo Services ASA||15628||0||1|
|Identive Group, Inc.||15658||0||1|
|Euronet Worldwide Inc.||15641||0||1|
|First Aviation Services Inc.||16005||0||1|
|Firestone Ventures Inc.||16001||0||1|
Thank you for your time.
11-20-2012 09:47 PM
We only have one example of a resignation in the data so this suggestion is interim:
First search for keywords such as President, Chief or Vice. If not found, return 0 for both columns
Probably exclude any row that has result in it
Next look for the string Resig. If found, assume it is a resignation
Other wise assume it is an appointment
I would suggest you need thousands of rows of data to fine tune the algorithms
Richard in Oz
11-20-2012 09:51 PM
Your new dataset does not look like your narrative. None of the sample announcements were for CEO. Also, Molex did name a new CFO, why was that not included.
Beyond those details, I am assuming that you do NOT have SAS text miner licensed. It is designed to answer just this sort of question.
Without it, you are going to have to write a lot of fairly tedious if-then-else logic in a DATA step to heuristically pull these codes out, and every time you get a new set of data, you will have to review the code.