Hewlett-Packard utilizes Text Mining
What is the practical Application of text mining?
The main practical application of text mining in this case was CRM focused and dealt with the process of how HP employed SAS Text Miner to determine what its most valuable customers were discussing and then how to develop new customer service strategies based on the recurring themes extracted from text mining. In HP’s case they wanted to monitor public insights and opinions (i.e. review sites, blogs) and bring science to customer relationships.
How do you think text mining techniques could be used in other businesses?
Measuring customer preferences - analyzing qualitative interviews
Fraud detection - investigating notification of claims
Improving search engines- spiders, keywords, etc…
Marketing Surveys – Know your customers
Email Clients – spam filters
Financial Services – for past, current and predictive modeling
Research and Development - Make what the customer wants
What were HP's challenges in text mining? How were they overcome?
HP realized they had a wealth of knowledge collected in the form of emails, chat conversations, telephone calls and messages, word documents and other business related communications that they could tap into to identify critical business information.
Because their initial system could not identify the relationship between text communications from their customer service center and key business trends, they turned to SAS Text Miner to create a hybrid mining system that utilized structured and unstructured data sets.
Text analysis is difficult because you have to quantify text as well as deal with other challenges such as dimensionality and data dispersal. The real challenge for HP was combining structured data with unstructured data. HP was able to utilize SAS Text Miner’s technique called singular value decomposition.
Furthermore the SAS analysts faced adversities trying to monitor and extract data from customer’s activity on the HP web site as well as
The main practical application of text mining in this case was CRM focused and dealt with the process of how HP employed SAS Text Miner to determine what its most valuable customers were discussing and then how to develop new customer service strategies based on the recurring themes extracted from text mining. In HP’s case they wanted to monitor public insights and opinions (i.e. review sites, blogs) and bring science to customer relationships.
How do you think text mining techniques could be used in other businesses?
Measuring customer preferences - analyzing qualitative interviews
Fraud detection - investigating notification of claims
Improving search engines- spiders, keywords, etc…
Marketing Surveys – Know your customers
Email Clients – spam filters
Financial Services – for past, current and predictive modeling
Research and Development - Make what the customer wants
What were HP's challenges in text mining? How were they overcome?
HP realized they had a wealth of knowledge collected in the form of emails, chat conversations, telephone calls and messages, word documents and other business related communications that they could tap into to identify critical business information.
Because their initial system could not identify the relationship between text communications from their customer service center and key business trends, they turned to SAS Text Miner to create a hybrid mining system that utilized structured and unstructured data sets.
Text analysis is difficult because you have to quantify text as well as deal with other challenges such as dimensionality and data dispersal. The real challenge for HP was combining structured data with unstructured data. HP was able to utilize SAS Text Miner’s technique called singular value decomposition.
Furthermore the SAS analysts faced adversities trying to monitor and extract data from customer’s activity on the HP web site as well as