Unearthing the Value of Text

Organisations collect huge volumes of information every day, around 80% of it unstructured text: customer feedback, e-mails, Web documents, memos, warranty claims, surveys, journal articles, research studies, CVs, competitive intelligence… the list is endless. But the information is under-exploited.

Nobody has time to read every source of information, much less the time or inclination to sort, classify and digest it. Because structured data could not be analysed effectively, nuggets of valuable information have remained buried, so important decisions end up being taken without the decision-maker having as full a picture as possible. SAS Text Miner has been developed to surface those hidden nuggets

Organisational performance is enhanced because SAS Text Miner enables the classification of multiple structured and unstructured fields. SAS Text Miner can dramatically enhance performance by distilling information from multiple business units into a format that's easy to manage and analyse. At the same time, a full range of predictive modelling tools mean you can better recognise trends and predict business opportunities.

• Universal data access and web crawling capabilities
• Extract, transform and load textual data into a SAS data set for mining
• Support for multiple languages
• Easy-to-use self-documenting interface
• Comprehensive text pre-processing capabilities
• Extensive feature extraction
• Dimension reduction techniques
• Text clustering algorithms
• Interactive results window
• Concept linking
• Hierarchical clustering browser
• Document categorisation
• Flexible reporting

Text mining in action
Text mining has already been deployed at SAS customer sites across the world. In the UK, for example, the Universities and Colleges Admissions Service (UCAS) relies on SAS to help it make accurate assessments of which students most deserve placement in undergraduate courses. In the US, American Honda has used text mining to improve safety and customer satisfaction while saving millions in warranty costs, while in Italy the City of Turin is using text mining approaches to monitor public opinion about the 2006 Winter Olympics and then target its communications campaigns.

Read text mining customer success stories at http://www.sas.com/success/tm.html