The emoticon analysis is a capability that comes from IBM's improvements in its text analysis. It is intended to prove better analysis to text information that often is devoid of context.

Emoticons have become part of our written language in email, SMS messages, blog comments and through games on set top boxes. The emoticons cut across international borders, providing a level of analysis for context across multiple geographic markets.

IBM uses natural language processing and adds other dimensions that can come from the emotions conveyed through the emoticons.

Additionally, the engine includes terminologies and 180 taxonomies related to an industry. This can help with differentiating brand names. For instance, orange can be the fruit or the telecommunications firm.

The analytics software pulls much of its data from the cloud even though it is on-premise technology. The next step is to provide the service in the cloud.

In the meantime, IBM customers are still a bit reticent about using the cloud for data warehousing purposes.

That's a practice that will have to change. The cost structure does not support the continued use of on-premise applications when the cloud can perform the same service at a lower cost and potentially better performance.