RoutePlanner Model – Support for Financial Brokers Using Visualization to Enhance Insight into Information

This paper considers the needs of financial brokers seeking to use information visualization to transform raw data in a data warehouse into enhanced insight, so as to identify marketing opportunities and better inform their clients. We introduce a model of the visualization process, called RoutePlanner, which was developed from a theoretical perspective. RoutePlanner comprises guidelines and stages for the identification of the data to be visualized, the selection of appropriate visualization methods, and the evaluation of the visualizations. In a collaborative process with a data expert and a visualization expert, the domain experts (brokers) were oriented in the use of RoutePlanner. They evaluated each stage of RoutePlanner positively, as well as the visualization process as a whole. The selected visualizations were successful in terms of the enhanced insight they provided, and the brokers who participated in the study were able to better inform and advise their clients in financial planning sessions. Key terms: Domain analysis, evaluation, guidelines, identification, selection, visualization methods, information visualization. Introduction This paper describes a study conducted in the application area of financial brokerage, where the focus area is the process of information visualization. The research, which originates from a masters degree study by the primary author, relates to ways of supporting financial brokers in using data warehousing and information visualization so that the brokers can be better informed and, in turn, also inform their clients. We present a model of the visualization process that identifies relevant data, selects and uses visualization methods to transform raw data in a warehouse to representations, and then evaluates the visualizations that provide enhanced insight, using guidelines for each stage of the process. The stages of the visualization process are described and the findings are given of surveys among the financial brokers both before and after exposure to the visualization process. Background Some key concepts of the study are described, after which we explain the context of the study in terms of the realworld problem and the target group. Material published as part of this publication, either on-line or in print, is copyrighted by the Informing Science Institute. Permission to make digital or paper copy of part or all of these works for personal or classroom use is granted without fee provided that the copies are not made or distributed for profit or commercial advantage AND that copies 1) bear this notice in full and 2) give the full citation on the first page. It is permissible to abstract these works so long as credit is given. To copy in all other cases or to republish or to post on a server or to redistribute to lists requires specific permission and payment of a fee. Contact [email protected] to request redistribution permission. RoutePlanner Model – Support for financial brokers using visualization for information gain 158 Data Warehousing According to a formal definition by Kimball (1996), a data warehouse is a repository of temporal, multi-dimensional, multi-variant, structured and abstract data, ordered in terms of facts and dimensions. He advocates the use of warehouses to represent organizational data. The data must be meaningful and interpretable and should provide the user with enhanced insight. Data warehouses store large volumes of data, with further data being added at regular intervals. Whereas a database contains only the most recent values, a data warehouse contains multiple data snapshots, thus allowing for the analysis of data trends and historical data. There are two approaches to the analysis of warehouse data, namely the automated and the manual approach. The automated option uses data mining algorithms to extract information. The manual option, which is the approach adopted in the present study, requires holistic userexploration and interpretation of the data. This exploration process can be guided and enhanced by visual presentation of the data, so as to discover data patterns, outlier values and data clusters. These findings, in turn, must be used to generate meaningful information that can be applied in the real-world context. Visualization Information visualization (IV) is the use of computer-supported interactive visual representations of abstract data to amplify cognition (Card, Shneiderman & MacKinlay, 1999). The purpose of visualization is the exploration of data and the presentation of information so that data owners can better understand their data and use any volume of raw data, including large volumes (Ward & Theroux, 1997; Wiss, Carr & Jonsson, 1998). In short, according to Saraiya, North and Duca (2005), the main purpose of visualization is the generation of insight. Relationship Between Data Warehousing and Visualization Information visualization, as opposed to scientific visualization, aims to visualize abstract data with no inherent structure. Such data may have no intuitive or formal visual representation, and IV is therefore an appropriate strategy for its exploration. A large collection of visualization methods (VMs) is available, with different methods serving varying functions. Shneiderman (1996) refers to IV in terms of seven abstract tasks that are included in his visual ‘information-seeking mantra’, namely: overview, zoom, filter, details-on-demand, relate, history and extract. Real-World Problem Under the present system, financial brokers in the target group use a financial needs analysis (FNA) program to access and to manipulate a database containing the most recent financial data for each of the clients in his/her client portfolio. The overview and analysis process may be conducted by considering a client’s resources and requirements, usually during a client-broker consultation, where the broker selects that specific client’s information from the portfolio, extracts a particular unit of data for analysis (e.g. pension, estate management or financial planning) and then views the client’s details for that unit. For the broker to view the data of all clients in a portfolio, a limited number of fixed queries are included in the FNA program, the results of which are presented as lists. The database does not store historical data, thus the broker sees only the data relating to the current status of each client. Thus, there are major limitations on the insights gained from the data in its present format and structure, using the current FNA program for representation of results.