Sessions
Session Abstracts
The following sessions will take place at Analytics 2013. This page is updated often so please check back frequently for the most up-to-date information.Social Network Analytics for Fraud Detection: Insights and Challenges
Data mining algorithms are focused on finding frequently occurring patterns in historical data. These techniques are useful in many domains, but for fraud detection it is exactly the opposite. Rather than being a pattern repeatedly popping up in a data set, fraud is an uncommon, well-considered, imperceptibly concealed, time-evolving and often carefully organized crime that appears in many types and forms. As traditional techniques often fail to identify fraudulent behavior, social network analytics offers new insights in the propagation of fraud through a network. Indeed, fraud is usually not something an individual would commit by himself, but is often organized by groups of people loosely connected to each other. The use of networked data in fraud detection becomes increasingly important to uncover fraudulent patterns and to detect in real time when certain processes show some characteristics of irregular activities. Although many analyses typically focus in the first place on fraud detection, the
emphasis should shift toward fraud prevention - detecting fraud before it is even committed. As fraud is a time-evolving phenomenon, social network algorithms succeed to keep ahead of new types of fraud and to adapt to changing environments and surrounding effects.
Revolutionizing Decision Making: How Analytics Will Take Over the Business
With advances in big data, artificial intelligence and increased metric captures of everything we do, analytics will go through a radical transformation in the next few decades. As a result, there will be a shift from analytics simply influencing business decision makers to analytics actually owning the decisions. This transformation is already happening in the pricing community, but expect it to expand to even CFO-level decisions.
Text Analytics and Latent Semantic Dimensionality
The emergence of big data analytics has generated a lot of interest in the quantitative analysis of unstructured text data. Customer comments, news stories, industry report segments, tweets and email messages are now routinely analyzed by text mining software solutions such as SASŪ Text Miner. Latent semantic analysis, a text analytic framework for extracting conceptual dimensions, offers solutions for analytic needs, including include document summarization and incorporation of unstructured text into quantitative predictive modeling. This presentation addresses the problem of latent semantic dimensionality selection. From simple visual examination of eigenvalue scree plots, to the implementation of an algorithm for multiple elbow point detection, the presentation will cover the detection of multiple dimensionalities in textual data. A number of illustration examples will show how document collections - including responses to open-ended surveys, customer comments and industry report paragraphs - can be
examined at alternative levels of semantic abstraction that represent topics, megatopics and microtopics.
Speech Analytics Applications to Predictive Modeling
Speech or voice analytics is an emerging technology that is gaining the interest of contact center operators, as it provides insights into a previously untapped, yet massive, source of information, which is the true "voice of the customer," i.e., the recorded phone calls. The potential range of applications of such technology is impressive: QA automation, 100 percent compliance adherence, call center metric (CSAT, AHT, Conversion, etc.) optimization through driver discovery, desktop performance optimization through targeted agent coaching, call model assessment, and predictive modeling. In this paper, we focus on the speech analytics technology application to improve the lift of propensity-to-pay models in an industrial setting by incorporating dynamic payment behavior indicators into existing static models. This approach is similar to using other types of unstructured data, yet quite different, in the sense that no text mining is performed on call recording transcriptions. The indicators are derived by the
speech analytics tool by analyzing combinations of phrases and behaviors without the use of transcription. This proposed solution optimizes call center costs while concentrating agents' efforts on the most lucrative accounts and addresses the challenge of static treatment plans by listening to customer responses. Results pertaining to expected lift are provided.
Predictive Analytics in Social Media and Online Display Advertising
The last decade has seen unprecedented growth in the space of online advertising and digital media marketing. The new wave of social media (Facebook, Twitter, etc.) is making it easier than ever for marketers to reach the right customers at the right time with the right products and offers. However, the marketers, online advertising platforms and other stakeholders need to be equipped with suitable analytical tools and methodologies to maximize the potential of online and digital media. The traditional analytical tools are often insufficient due to the rapidly growing volumes of data as well as the increasing importance of dealing with textual and unstructured data. In this talk, we will present three case studies on applying data analytics in social media and online display advertising to help our clients stay competitive in the marketplace.
