The founders of Complexica have worked together for almost 20 years building enterprise software applications that are “smart” and “adaptive”. Such applications bring together the disciplines of algorithmic science and software engineering, as the “science” enables the "smarts" within the each application. By applying decades of research in automated data mining, analytics, and expert systems, Complexica has developed Larry, the Digital Analyst® – a cloud-based software application that incorporates the latest advances from the field of Artificial Intelligence called Cognitive Analytics.
In general terms, Cognitive Analytics refers to sophisticated algorithms that can manage and analyse big data sets to help people make better, faster decisions. Each day, enterprises collect or aggregate vast amounts of data from very diverse sources and want to analyse these data sets to gain specific insights – for example, to understand how different customer segments respond to different product and service offers, how customer preferences change over time, how customer loyalty is generated and maintained, how promotional campaigns impact volume, margin, and product cannibalization, and so on. The accuracy of such insights can provide a competitive edge.
Based on Cognitive Analytics, the primary task of Larry, the Digital Analyst® is to manage the interaction with internal data warehouses and external data sets, and to automate the associated data loading, handling, and analysis processes. Larry consists of a collection of machine-learning algorithms (e.g. Bayes nets, artificial neural networks, rough sets, classifier systems, support-vector machines, decision trees, genetic algorithms) that are used for various data mining problems and advanced analytics. These include:
- Anomaly detection (outlier/change/deviation detection) is used to identify unusual data records that might be interesting or contain data errors that require further investigation. Anomaly detection is of key importance in the process of cleaning (and understanding) data.
- Association rule learning (dependency modelling) is used for finding relationships between variables. For example, a supermarket might gather data on customer purchasing habits, and using association rule learning, the supermarket can determine which products are frequently bought together and use this information for marketing purposes. This is sometimes referred to as market basket analysis.
- Clustering is used for discovering groups and structures in the data that are in "similar", without using known structures in the data. Data is viewed as points in a multidimensional space and points that are “close” in this space are assigned to the same cluster. For example, an organization may group their customers into micro-clusters for personalized promotional campaigns.
- Classification is used for classifying new data to existing categories. For example, an expert system might attempt to classify a credit card transaction into a "legitimate" or "fraudulent" category.
- Regression is used to find a mathematical function that models the data with the least error. In other words, regression is a statistical process for estimating the relationships among variables (i.e. the relationship between a dependent variable and one or more independent variables). The discovered relationships aide in understanding how the typical value of the dependent variable changes when any one of the independent variables is varied. A common application of regression is in the area of price elasticity.
- Summarization is used to provide a more compact representation of the data set, including visualization and report generation. For example, after clustering is carried out, the clusters themselves are summarized (e.g. by generating the centroid of the cluster and the average distance from the centroid of points in the cluster) and these cluster summaries become the summary of the entire data set.
Larry, the Digital Analyst® is deployed through our sales optimisation software to address a variety of problem domains, such as increasing the sales effectiveness of telesales and in-field reps, maximizing customer loyalty and engagement, capturing new revenue and margin opportunities, influencing customer buying behaviour and preferences, understanding price elasticity – among many others. Because most companies maintain large data warehouses on what their customers bought and when, it’s possible to apply Larry to these problem domains to automate the data analysis process and create actionable insights that generate real value.