Swiss CRM Forum
I recently went to the Swiss CRM Forum (Swiss Customer Relationship Management Forum) in Zürich. The topics where all related to CRM data with, for example, data management, business intelligence and data analytics. Among other, there were people from SPSS, SAS, Teradata, SAP, Microsoft, etc. It was very interesting to discuss with these people about future trends of software such as PAWS (formerly Clementine), from SPSS and Customer Experience Analytics, from SAS.
Unfortunately, most of the presentations were in German. It seems the french part of Switzerland was not the main audience 
However, I’ve attended to a very interesting presentation from eBay. This one was by chance in english. The presenter explained how eBay was using the keywords of incoming visitors to personalize their homepage. In other words, when you reach www.ebay.com, the content of the page may depends from the keywords you have typed in Google (if you come from Google). This is clearly one of the main aspect of the Web 3.0: web recommendation and personalization.
Link to the Swiss CRM Forum.
Small Book Review: The Little SAS Book
To start learning SAS, I decided to read a book different from the dozens of SAS users manual in PDF. According to the nice Amazon reviews, I took the one from Delwiche and Slaughter: The Little SAS Book, A Programming Approach (4th ed.). The book is easily readable since it is composed of two-pages articles. Each one focuses on a specific task or function of SAS. The authors discuss about SAS Base programming notion (no Data Integration Studio or Enterprise Guide GUI). Note that they also wrote a book for Enterprise Guide.
The book is divided in 10 chapters that go through reading data sets, building reports, combining data sets, writing macros, using grahics, debugging SAS programs, etc. Some appendices are very useful such as “Coming to SAS from SPSS” and “Coming to SAS from SQL”. The last one is important since the SAS SQL langugage (PROC SQL) is slightly different from the standard SQL langugage. After working on SAS with this book for a few months I realised that it is very convenient for basic tasks. Any simple task you don’t know? Just look in the index and you will find the corresponding function. However, for more advanced topics, the book is a bit light. Well, according to the title, this seems to be normal.
The Little SAS Book is very easy to read. In addition to being easily searchable, one can read it from A to Z. Although technical, it is not boring since subjects are divided in two pages. In conclusion, the Little SAS Book is an excellent book to start SAS with. It is also a very useful companion book for basic functions. For more advanced issue, you will have to look in the SUGI papers, ask the SAS support team or find a more advanced book. To be honest I liked the book so much that I ordered the Enterprise Guide one…
Privacy in Behavioral Targeting: Poll Results
Here are the results from the poll on the privacy apects of behavioral targeting. Thanks to the 27 people who participated. The question was: “How do you feel about behavioral targeting in online advertising?“:
- I don’t care (7%)
- I want to be informed about the use of my profile data (11%)
- I want to be informed about the use and to be able to access my profile data (15%)
- I want to be able to deny behavioral targeting (and delete my profile if existing) (67%)
Of course the results are not statistically significant but they still give an indication of the different opinions on the subject. Personally I voted for the second answer (of course, I’m not objective since I work in the field of behavioral targeting). I’m quite surprised by the high percentage for the last answer, which is the most conservative one regarding our personal data. People who voted for the last answer should maybe reconsider their use of tools such as LinkedIn and Picasa, since both keep personal data even when the account is closed.
The GenIQ Model Modeling and Data Mining Software
Here are a few words on the GenIQ model, from Bruce Ratner.
The GenIQ Model is a machine learning alternative model to the statistical ordinary least squares and logistic regression models. GenIQ lets the data define the model – automatically data mines for new variables, performs variable selection, and then specifies the model equation – so as to “optimize the decile table,” to fill the upper deciles with as much profit/many responses as possible. Put differently, GenIQ seeks to maximize cum lift, a measure of model predictiveness of identifying the upper performing individuals often displayed in a decile table. GenIQ produces models that outdo statistical models, and is a different model: unsuspected equation, ungainly interpretation, and easy implementation.
Database Marketing (DM) regression models seek to maximize cum lift, a measure of model predictiveness of identifying the upper performing individuals often displayed in a decile table. DM regression models built on today’s big data – consisting of a multitude of variables, an army of observations – using statistical regression models, conceived and testing within the small-data setting of the day, 205 years ago, is problematic: Fitting big data to a pre-specified small-framed model produces a skewed model with doubtful interpretability and questionable results. The GenIQ Model is a machine-learning alternative regression model to the statistical models. It is an assumption-free, free-form model that maximizes cum lift, equivalently, the decile table. Sign-up for a free GenIQ webcast: Click here.
Data Mining Poll: Online Privacy
Since I’m applying behavioral targeting in online advertising, I’m curious to know what you think about privacy issues of such ads. Often, laws are well defined for data protection but not specifically in the online world. With the following poll, I would to know your opinion about the privacy when surfing on a website with advertisement that are based on your surfing profile:
Feel free to comment if your choice is not in the box or if you would like to add some remark.
-
Data Mining Research
Sandro Saitta on AnalyticBridge, LinkedIn, Twitter, Delicious and Technorati.
Join the Bloggers group! -
Visitors
-
Networks
-
T-shirts, Mugs & Mousepads
