Mining Unstructured Data

The information entered as “tweats” by a lot of people inside applications such as Twitter, Linkedin is unstructured. The “tweats” updated into such applications are similar to our own thought processes. Data Mining techniques involve mining on data that is precisely defined. For example a product survey contains questions such as Which color do you like the most? Which feature do you like the most, so on and so forth.

By writing some standard OLAP processing logic one would be able to derive reports required for providing critical business intelligence reports. In this case, there is also a considerable amount of effort spent on data definition, data entry and data analysis.

Tweats contain a lot of unstructured information. Rather than setting up a review committee to provide reviews on movies, products, packaged food,services so on and so forth, one can poke or construct a review system based on information updated into Twitter, Mouth Shut.com, Linkedin, facebook etc.. The challenges obviously would be to construct a mining system that would be based both on likelihood and statistics.

The user responses or tweats will be mapped with certain possible values. For example a tweat such as “Oh great I had a good time at the coffee shop” could indicate any value between 7 to 10 on a rating system. In the case of consolidating unstructured information, statistical inferences will be combined with likelihood. So the same tweat can be used to infer two similar situations or view points.

There are already some applications that provide Product reviews based on tweats inside Twitter. It is now for the developers to develop some more applications that can effectively consolidate responses in the form of tweats and also derive business intelligence.

Continue Reading

10 Very Interesting People (VIP) in Data Mining

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).

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).

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).

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).

For More Information about Data Minining click here

Continue Reading

What’s Your Excuse For Not Using Data Mining?

In an earlier article I briefly described how data mining can help marketers be more efficient (read… increased marketing ROI!). These marketing analytics tools can significantly help with all direct marketing efforts (multichannel campaign management efforts using direct mail, email and call center) and some interactive marketing efforts as well. So, why aren’t all companies using it today? Well, typically it comes down to a lack of data and/or statistical expertise. Even if you don’t have data mining expertise, YOU can benefit from data mining by using a consultant. With that in mind, let’s tackle the first problem — collecting and developing the data that is useful for data mining.

The most important data to collect for data mining include:

Transaction data – For every sale, you at least need to know the product and the amount and date of the purchase.
Past campaign response data – For every campaign you’ve run, you need to identify who responded and who didn’t. You may need to use direct and indirect response attribution.
Geo-demographic data – This is optional, but you may want to append your customer file/database with consumer overlay data from companies like Acxiom.
Lifestyle data – This is also an optional append of indicators of socio-economic lifestyle that are developed by companies like Claritas. All of the above data may or may not exist in the same data source. Some companies have a single holistic view of the customer in a database and some don’t. If you don’t, you’ll have to make sure all data sources that contain customer data have the same customer ID/key. That way, all of the needed data can be brought together for data mining.
How much data do you need for data mining? You’ll hear many different answers, but I like to have at least 15,000 customer records to have confidence in my results.

Once you have the data, you need to massage it to get it ready to be “baked” by your data mining application. Some data mining applications will automatically do this for you. It’s like a bread machine where you put in all the ingredients — they automatically get mixed, the bread rises, bakes, and is ready for consumption! Some notable companies that do this include KXEN, SAS, and SPSS. Even if you take the automated approach, it’s helpful to understand what kinds of things are done to the data prior to model building.

Preparation includes:

Missing data analysis. What fields have missing values? Should you fill in the missing values? If so, what values do you use? Should the field be used at all?
Outlier detection. Is “33 children in a household” extreme? Probably – and consequently this value should be adjusted to perhaps the average or maximum number of children in your customer’s households.
Transformations and standardizations. When various fields have vastly different ranges (e.g., number of children per household and income), it’s often helpful to standardize or normalize your data to get better results. It’s also useful to transform data to get better predictive relationships. For instance, it’s common to transform monetary variables by using their natural logs.
Binning Data. Binning continuous variables is an approach that can help with noisy data. It is also required by some data mining algorithms.

Continue Reading

Standard Process for Data Mining?

What if you’re assigned to a new project that involved analysing large data (e.g. transaction database, market or patient data) for your organization. What you should start first? Well, there are many standard process you can begin with but there is one in specific developed for industry called The CRISP-DM project. Starting from the embryonic knowledge discovery processes used in early data mining projects and responding directly to user requirements, this project defined and validated a data mining process that is applicable in diverse industry sectors. This methodology makes large data mining projects faster, cheaper, more reliable and more manageable. Good luck with your project then.

Continue Reading