Artificial intelligence and data warehousing

Artificial Intelligence & Data Warehouses

Artificial Intelligence and Data Warehousing

As the complexity and structure of databases grows as a result of enterprise-class data warehouses being able to interpret and use unstructured content (Doan, Halevy, 2005) there is a corresponding need for building linguistic models out of semantic data using artificial intelligence (AI) (Fortuna, Mohorcic, 2009). The techniques of latent semantic indexing (LSI) and contextual factor analysis are increasingly being used to find patterns in unstructured data within data warehouses (Doan, Halevy, 2005). Artificial Intelligence is also found in constraint-based approaches to defining complex Structured Query Language (SQL) systems as the path to complete queries can be optimized over time. All of these examples illustrate how critical AI is today in the managing of data warehouses that are comprised of multiple sources of data, some structured and some unstructured. The use of constraint-based modeling in conjunction with semantic query and modeling will significantly streamline the use of data warehouses, yielding insights not possible from only working with structured data.

Just as AI is significantly changing how data warehousing is being used as a strategic resource in companies, it is also fundamentally changing the nature of business today. Most significantly, the use of LSI technologies to create more effective insights into how to improve customer service as evidenced by the use of AI was part of Decision Support Systems (DSS) (Phillips-Wren, Mora, Forgionne, Gupta, 2009) is growing. Second, the creation of ontological databases that aligns to person's roles (Pinto, Marques, Santos, 2009) is also now possible. This translates into the use of AI to provide contextual guidance to decision makers based on their previous preferences and previous patterns of decisions, making DSS results optimized (Phillips-Wren, Mora, Forgionne, Gupta, 2009).