Business Intelligence Lowdown

Microsoft Plans to Take-on Google on the Search Turf

When I want to search for something my fingers seem to automatically type in www.google.com into the browser window. It not only comes spontaneously to me, as I am sure it does to millions of Internet users worldwide, but also seems the most appropriate place to start the search.

Microsoft Plans: The recent report by the European President of Microsoft Corp. that it plans to introduce a search engine that's better than Google in six months in the US and Britain only indicates the domination Google enjoys today. Microsoft believes being good is not enough to win the hearts and minds of consumers already dedicated to another standard. The Corporation feels the way ahead is to make its search engine available in a place where it will be most widely used like MSN and Hotmail.

Search Truths: How many of us really utilize the search to its fullest? Have you ever used a search engine to locate a restaurant close to your house that sells fish and chips on the menu for under $5 a plate or perhaps locate the best shopping deals on the net? Generally these days you get back URLs based upon research and chances are that 50% of the time you do a search and you don't get the URL you're looking for.

The Future: The search of the future has not only got to be fast and exhaustive but also specific to what the user is looking for. It’s a bit like going into shopping site and searching the site for a gift for under $50 and being presented with a web page filled with choices. I should be able to get the information I am looking for from the very web page I am in without being directed to another website or another URL

Internet Giants realize that they will have to better their search technology if they have to be in the run tomorrow. Microsoft’s recent move is just one small step. After all, if Google is here can Microsoft be far behind?

Data mining - A BI threat to Security?

"Data mining" - is associated with Business Intelligence and it helps gather information. Unknown to many an internet user, it is being used by several companies to learn about who you are and what you do. Initially they were introduced as personalized pages.

Sure it looked good to have a web page with a layout you had set when you last logged in - the same colors, themes and contents. It is also amusing to have a small window pop up which says "Hello James... looking for a house loan". But, how on earth did they know that you were looking for a loan or a house or a car?

Data mining is also a concern with governments eavesdropping on citizens. Computer technology is used to extract knowledge that's buried in enormous volumes of undigested information. With the speed at which computers can crunch data today searching through a few hundreds of thousands of lines of data to filter out a few key words - it is easy.

So if you searched for “Hijack” and related words…its easy to pick you up. Similarly the contents of your chats can be searched in no time to detect possible pedophiles and murderers or even tax evaders. While these are the high points of Data Mining, it can also be used to search and sort through trillions of bits of raw data culled from telephone calls, e-mails, the Internet, airlines, car rentals, stores, credit card records and other sources.

To a common man are these not the very threats he was trying to shield himself from by using a firewall and anti spy software and anti virus software. And here reputed companies and makers of search engines are doing the same thing. What’s more worrying is that these are the very useful sites you just cannot do without. So in a creepy way even the innocent citizen has been given the same"preferential" treatment as Bin Laden. Now is that not a misuse of technology and invasion to one's privacy?

Hybrid OLAP Tools

Online analytical processing (OLAP) tools are now using visual presentations like charts, maps and diagrams to help users detect data patterns that help in the decision-making process. A few vendors have come out with innovative implementations of the OLAP technology.

• Tableau from Tableau Software generates tables from standard structured OLAP and relational data sources, has an interactive display, and can hold embedded charts with multidimensional data views within cells.
• Kheops Technologies' JMap Spatial OLAP extension presents a multidimensional view of data through interactive maps, tables and diagrams. It also offers Web map navigation and manipulation with a modular architecture.
• Text OLAP that is a part of Megaputer's PolyAnalyst suite of data and text mining tools delivers an interactive, structured analysis of text documents. The difference between conventional text mining is evident in the use of extracted terms and concepts as dimensions. Entities are displayed in the matrix form with linked charts and color-coded extracts from the reference documents.

Fighting Future Crime

The utility of business intelligence (BI) is spreading across all fields; BI software is being deployed for a variety of purposes ranging from preventing theft to deterring crime. The Richmond Police Department (RPD) in Virginia is implementing data mining, predictive analysis, and BI tools from enterprise business intelligence software solutions providers SPSS and Information Builders to prevent future crime from happening. The application, which combines SPSS' predictive analysis and Information Builders' enterprise BI capabilities along with an analytic framework from RTI International, an independent scientific research and technology development corporation, will aid the RPD in deploying police power to the right places at the right time and thus preventing crime. The software and implementation make sense since criminal activity follows predictable patterns, and the police are natural data miners who excel at wading through data to derive solutions to crimes, says Colleen McCue, senior research scientist at RTI International. Eweek reports:

"By better characterizing crime trends and patterns to predict where crime was likely to occur, the police department could anticipate criminal activity and proactively place their resources," McCue added.

Decision Support Systems

Decision Support Systems (DSS) are interactive, computerized information systems that lend support to the decision making process. There are five specific types of decision support systems:

1. Communications-driven: A DSS that emphasizes communications, collaboration and shared decision-making support.

2. Data-driven: A DSS that emphasizes access to and manipulation of a time-series of internal company data and occasionally, external data.

3. Document-driven: A relatively new form of DSS that is focused on the retrieval and management of unstructured documents.

4. Knowledge-driven: A DSS that suggests or recommends actions to managers through expertise or knowledge about a particular domain, understanding of problems within that domain, and skill at solving some of these problems.

5. Model-driven: A DSS that emphasizes access to and manipulation of statistical, financial, optimization and/or simulation models.

Data Mining Helps Database Marketing

With growing emphasis on the need for maintaining the quality and integrity of data, organizations are spending their time, resources and money in cleaning up their databases. This cleanup process involves replacing outdated data and consolidating data from various sources. But in enterprises that employ database marketing, the purging process can often be leveraged to segment the market in such a way that the most-likely-to-purchase customers can be targeted to help increase profits.

Data mining can be used to the company's advantage if combined with effective market segmentation techniques. Instead of the brute force approach in which all customers are targeted in a marketing campaign, segmentation techniques identify those clients who will perform the best, and thus ensure profitable returns on the campaign. Market segmentation intelligence should be applied to the database to add value to each of the customer records, which can then be used to prioritize customers and rank them according to their order of importance.

Some organizations use a segmentation technique that links multiple customer and prospect databases in a relational manner. These test segments or "cubes" are then tested using discriminating identifiers to decide which segments are of most value to the enterprise. Market segmenting intelligence thus gained can be utilized to build future marketing programs for bigger databases, which in turn will lead to higher returns for the organization, within the given marketing budgets. The bottom line is that sales increase in a timely manner at a lower cost.

Multilingual Text Mining Applications

Text mining solutions providers, TEMIS and Basic Technology, have teamed up to develop new multilingual text mining applications for government and commercial markets. TEMIS’ Insight Discoverer solutions for European Languages and Basis Technology’s Rosette language analysis system for Asian and Middle Eastern languages will integrate to aid users in extracting meaningful information from unstructured data. KM World reports:

Basis Technology will use its expertise in linguistics products to ensure technical assistance, support, maintenance and further development of TEMIS text mining products for that company's North American customers. The companies plan to collaborate in the future on text mining solutions for vertical markets such as life sciences, financial, legal discovery and CRM.

Data Mining Software from IBM

The DB2 Content Manager Standard Edition v8.3 for iSeries is the latest data mining software from IBM, intended for use in small to medium-sized medical offices, law establishments, retail outlets and manufacturing organizations to help them consolidate unstructured data into organized and meaningful structures. This will in turn simplify the querying and reporting processes. Though the software is specifically designed for the iSeries market, it is not bundled with any hardware systems, according to Theresa O'Neil, director of content management software at IBM. Search CRM reports:

The content manager features a graphical workflow builder, a tool that allows users to manage document processes, such as document creation, approval and archiving. The DB2 Content Manager Standard Edition v8.3 for iSeries also offers Web services support, automated management of XML documents and XML schema mapping utilities.

Bringing Down Search Time

With increased demand for software that can cut both the time and cost involved in searching through volumes of email for a particular message, more and more companies are joining the list of such providers.  The latest to jump on the bandwagon is the newly launched Santa Clara-based Clearwell Inc., which rolled out its Clearwell Email Intelligence Platform earlier this week. Firms estimate that a typical request for email in a legal discovery situation requires 1,300 hours of labor and costs more than $100,000. Search tools are also available from the stables of other new companies like the New York-based Orchestria and the Washington-based MessageGate. Zdnet reports:

Bigger players like Computer Associates, EMC, HP, IBM, Symantec and Zantaz already offer some basic search capabilities in their email archiving programs.

SAS Ranked Among Leading Data Mining Vendors

Data mining software from SAS has been rated among the best in the industry by the research group Gartner Inc. The software vendor found itself in the top quarter of the “Magic Quadrant for Customer Data Mining, 1Q06” report, for excellence in the data mining segment and for significantly impacting the market growth and direction. The data mining solution, SAS Enterprise Miner, and the text mining solution, SAS Text Miner, were evaluated for the report. Business Intelligence reports:

SAS' award-winning data mining solutions also are available as part of SAS Customer Intelligence. The software enables organizations to find and retain profitable customers using demographic data, customer buying patterns and other data.