First, I have to admit that I really liked the book. It is nicely written and can be read from the beginning to the end like a story. The book is divided in two main parts. In the first part, the authors describe the DELTA methodology. DELTA stands for Data, Enterprise, Leadership, Targets, Analysts. Each of these words constitute a chapter in the book. The second part, Staying Analytical, is mainly about building an Analytical culture in your company.

Among interesting points, one can note a table about typical decision making errors. Also the five different analytical stages in which a company can be. An original example with actor Will Smith and how he is using Analytics in his work. All these are very interesting examples that make the book really valuable. One important message of the book is that any company can gain from analytics. The only question, that the book answers, is at which level.

In conclusion, this book is a must have for anybody involved in analytics, whether it is data mining, web analytics, business intelligence, etc. It contains several important advices for people involved in analytics within a company: how to manage project related to analytics, how to educate your colleagues about analytics, and so on. As written above, I will soon post an interview with Jeanne Harris in which you will learn more about Analytics at Work.

Several factors can influence the results obtained: the quality of the CRM data, the granularity of the web log aggregation, the data mining technique used, the size of the time window to build the ECP, etc. In this post, I will focus on the time window size. Without going into too much detail, there are two choices: short or long term time window. As example, short term can be from 1 to 7 days, while long term can be 7+ days. The choice of the time window size (short or long) is an important decision that will affect the results.

The choice depends on what we want to capture. If we want the recent interests of customers, then a short time window should be used. In this case, only the very recent web activities of costumers are taken into account. It will thus have a bigger variation over the time. In the case of a long time window, the overall customer interest is taken. In our project, we identify the customer at the house hold level (rather than at the person level). It is thus not possible to differentiate between the father and the son who are both using the same internet connection. Knowing this, getting the recent and volatile interest of the customer makes no sense and we have decided to use the long time window.

• Analytics at Work (Davenport/Harris): I recently had the chance to receive an advance reader’s copy of this book (publication date: February 12th). Comparing to Competing on Analytics, Analytics at Work is for a larger audience since it explains how any company can improve in analytics. This book is easy to read and contains a lot of advices for anybody involved in this field.

• Handbook of Statistical Analysis and Data Mining Applications (Nisbet/Elder/Miner): I heard a lot about this book. It seems to be THE new book about data mining. So I wanted to check by myself. I have to admit that it is a very interesting reading but since it is not theoretical, it is more beginner oriented. However, the applications and examples in SAS, Clementine and STATISTICA are very well presented. It is a new and fresh approach to data mining, which is very pleasant to read.

Even before making the final reviews, I can already recommend you these two books. Happy reading!

As a layman who is not familiar with technological terms and their meanings, if you come across the term “data mining”, it’s not surprising if you wonder where from and why data has to be mined. To explain it as simply as possible, data mining is the process of finding logical patterns in data and according order and meaning to various sets of seemingly random data. If you’re still in the dark and cannot understand this explanation, then perhaps you need to sign on with Google for its free email, because for those of you who have ever used any of Google’s free services, it’s easy to understand what data mining is all about.

When you open your Gmail account, there is a column to your right where a series of advertisements are displayed. They change every time you log in to your mailbox and every time you open a new email. And if you look closely at these sponsored links, you’ll find that they are based on the contents of your email. No, Google is not looking over your shoulders, reading your mail, and then directing ads to your inbox; rather, bots scan through your mail messages on the Google server, mine the data that is found there, and use keywords to direct the relevant sponsored links to your mailbox.

You can argue the privacy issue until you’re blue in the face, but Google is just not going to do anything about it, because as far as the search engine giant is concerned, there is no human eye reading your email other than your own. It’s all done automatically, for millions of users the world over, and the amazing aspect of this is the accuracy of the advertisements that are displayed on your screen.

They’re generated based on keywords in your email, and there are times when you feel that the Google data mining bots are able to think for themselves too – for example, if there is any reference to a native American name, there are advertisements relating to tourism relating to native American history. That’s how intuitive Google’s data mining process it.

At times it wows your mind, and at others, you feel that your privacy is being invaded, that someone is reading your mail, something that is meant for your eyes alone. And it’s not just your mail that is analyzed – if you’re logged on to Gmail or any other service from Google when you run a Google search, your search history too is used to generate advertisements. Your browsing habits are being analyzed by Google to generate the ads that you are most likely to be interested in.

Data mining is mostly used by marketers and people who need to analyze large volumes of data and make sense of it for some purpose or the other. And even though the example I’ve cited makes a shiver run down your spine, it’s the easiest way to explain it in non-technical terms.

To know more about Barbara Williams, visit her website: http://becomingacomputertechnician.com/

• Blog by bruno: This blog covers a very large number of topics including web data analysis and data visualization. Bruno also has an interesting list of data exploration tools.
• Datamining-blog: The focus of this blog is on data mining for CRM marketing. An interesting point about this blog is that it is written both in English and German.

I wish all the best to these blogs and I welcome them on the data mining list of blogs.

• Gregory Piatetsky: Author of the most popular newsletter in the data mining community, he has recently updated his website with new content. You can now subscribe with RSS and you can find KDnuggets on Twitter. Gregory does an amazing job in collecting data mining related information, analyzing it and distributing it to data miners (website, twitter).
• Bruce Ratner: He is author and his website contains several articles about data mining. He has recently been very active on social networks such as LinkedIn (website).
• Ajay Ohri: I think this is the most active blogger in the data mining field. He is very active on many social networks and has an excellent collection of interviews with key people in data mining and related fields (blog, twitter).
• Vincent Granville: As the creator of AnalyticBridge, Vincent has made a great job in building a community of people specialized in analytics fields. His network links more than 6600 members. So, it’s time to subscribe! (website, twitter).
• Matthew Hurst: He is the author of the very famous blog “Data Mining: Text Mining, Visualization and Social Media”. He is very active on his blog on topics such as social media and data mining the blogosphere (blog, twitter).
• Dean Abbott & Will Dwinnell: I put them together since they are co-bloggers. Abbott’s Analytics is an excellent blog (one of my favorite) related to data mining. When reading the posts, you can really feel the experience of the authors (blog).
• Greg Linden: His famous blog – Geeking with Greg – is well known for a while now. He writes very informative posts about personalization related topics (blog).
• Matt Cuts: He mainly writes about Google stuffs and SEO. However, he is also well know in the data mining world since several posts are directly or indirectly related to this field (blog).
• Themos Kalafatis: He writes a lot about text mining (social network mining, etc.) and his posts are very practical. It is always a pleasure to read his blog (blog, twitter).
• Randall Matignon: He is the author of very comprehensive books on SAS Enterprise Miner. You can find all information about his books on his webpage (website).
  

If you have been impressed by other people in the field of data mining, feel free to comment this post.

1. List of Data Mining Blogs: It is always very popular on DMR, the list of blogs now contains 43 data mining related blogs.

2. Data Miners on Twitter: It seems that data miners are quite active on Twitter… and that’s good news!

3. Data Mining Tools: From SAS to R/Java: The question of which language/tool to use for data mining is always important. Another important question is how to skip from one tool to another…

4. Five Reasons Why “Data Miner” is the Best Job in the World: It seems this 2nd degree post was appreciated by a lot of readers… and maybe future data miners

5. Data Mining Search Engine: With more than 8000 queries in 20 months, this data mining customized search engine is used every day.

You have maybe noted that DMR has a new logo. I hope you like it. I wish you a happy new year 2010 and I’m looking forward to discussing with you here, on Data Mining Research.

Sandro.

• Macro name: Get_data
• Input: myDataset (name of the dataset), myLine (line number), myColumn (name of the column), myMVar (name of the macro variable used to store the result)
• Output: A new macro variable named &myMVar.

And here is the corresponding code:

%MACRO Get_data(myDataset=,myLine=,myColumn=,myMVar=);
%GLOBAL &myMVar.;
data _null_;
set &myDataset.;
if _N_ = &myLine. then do;
call symput(symget('myMVar'),&myColumn.);
end;
run;
%MEND Get_data;

Of course, I would be very interested if someone found a simplier way of performing this action (and I think SAS support would be interested as well).

Today, Web Analytics (WA) are essential for companies active on the web. In Europe and especially in Switzerland, Web Analytics are a new field of activity but are necessary to leverage online communication and marketing. The potential is enormous and it is time to integrate Web Analytics in every communication and marketing strategy. But Web Analytics are not enough. Measuring and analysing your online data have to be integrated with offline data and analysis.

On January 27th, we will explore what will be the future of Web Analytics and their integration with offline (customer) data.

You want to share your experience of Web Analytics with others?
Discuss about your issues and your success with other professionals?
Or you just want to get in touch with people working in the WA field?

Take this opportunity and join us January 27th, 2010 for the first Web Analytics Wednesday of the year sponsored by HIPPARCOS and FinScore (www.finscore.com).

On the agenda (to be confirmed):

* Opening words
* Use case by practitioner (speaker tbc): Web Personalization
* FinScore: Online and offline become 1
* Discussion and networking

We are happy to welcome you at this free event.

Location: Forum Crédit Suisse – Rue de Lausanne 17 – 1201 Geneva (Only 1 minute walk from the main train station)
Time: 17h00

It’s now time to register!
http://www.webanalyticsdemystified.com/wednesday/list.asp?event_id=2995

If you have any questions or need details, feel free to contact us:
Sandro Saitta
FinScore S.A.
+41 (0)21 647 77 13
Contact form

1. We can predict the future
Yes, we can. Of course, this is not 100% sure, but still we predict the future. As long as we have enough data with some patterns inside (the famous gold nuggets), we can predict the future. What product will this customer buy next? What will be the pollen concentration in the air tomorrow? How will the stock market evolve in the following week? Yes, to some extend, and with some errors, we can predict the future. Isn’t that cool?

2. We can change our job without changing it
Sounds strange? Let me explain this. Data mining is a field where we apply machine learning and statistical techniques to some concrete problems in a certain field. Every new project may cover a different field. This gives you the opportunity to discover and learn new domains without changing your job. While working as a data miner, I have – in six years – discovered fields such as meteorology, civil engineering, finance and telco. Isn’t this great?

3. We can easily impress friends and family
This is not the case in every job. Being a data miner, it is. Of course, we could use simple terms such as “data, analysis and results” to explain what we do. However, we prefer to use terms such as “random set of points, support vector machines and score probabilities” instead. It is much more impressive, so why not use them? Yes, we are like this. Aren’t we awesome?

4. We can spy on anybody
Who has never dreamed of being a spy like James Bond? Being a data miner, we are kind of spies, at least according to the non-technical news. We steal data and we mine them to discover very personal information. We break the privacy of every human being on Earth by spying on them. Yes, we do that in our everyday job, and to be honest we love that. Of course, we could stop, but it is too much fun to play the spy. Isn’t it?

5. We never fail
Sounds impossible? It isn’t: we are data miners. Which means that if we find patterns in your data, we are good. If we don’t – it’s not our fault – it means there are some data quality issues (missing values, not enough data, etc.). But it’s not our fault. And if you insist, we can still find something for you. You bet? Give us any data (even random!) and we will find some patterns using clustering, for example. This is like finding patterns in clouds. If you look long enough, you will always see an image appear. Isn’t it beautiful?

If you think of other reasons, feel free to comment on this post and help me explain why “data miner” is the best job on Earth