MIT CFB
 

Summer 2008 Research

The MIT Center for Future Banking

(projects as of August 15th, 2008; i.e. about two months since students joined CFB)

Projects


A New Type of Savings Account

Leslie Gee, Ana Chen and Marko Popovic
(lgee4[AT]mail.rochester.edu, agchen[AT]MIT.EDU, marko__[AT]mit.edu)

Presentation http://web.media.mit.edu/~marko/ALL/GeeChenPopovic.ppt
(if you can not access this link please click on title)

Savings account holder sets predefined goals and shows financial and other progress update on the accompanying web based platform. Account holder invites family members and friends to contribute and provide support and advice through web based platform/blog. Monetary contributions will be posted either anonymously or publicly. This service could be linked with other bank services like the "Keep the Change" program. This service could improve financial literacy with younger customers. Potential benefits: new customers attracted to self-customized banking, increased investments in savings accounts, builds relationship and trust with younger customer. (a few demo frames http://wiki.media.mit.edu/pub/CFB/FinanceVisualization/combined.swf )


A New Type of Advanced Search Engine for Home buyers

Ana Chen, Mary Hong, Marko Popovic and Dan Ariely
(agchen[AT]MIT.EDU, maryh10[AT]MIT.EDU, marko__[AT]mit.edu, ariely[AT]MIT.EDU)

Presentation http://web.media.mit.edu/~marko/ALL/ChenHongArielyPopovic.ppt
(if you can not access this link please click on title)

When people make the wrong choices, it causes problems for them and the bank; delinquency and foreclosure rate for all mortgages was 7.3% in 2007, 5.7% of home equity lines of credit were delinquent or in default at the same time, in past 12 months more than half million US homes foreclosed and are now owned by the banks. Proposed web based intelligent search engine assists home buyers in finding the "right" house -realistically and financially, in a way that is logical, but natural. Proposed system a) considers buyer's present situation and future plans, b) allows buyer to accurately evaluate trade-offs between what he/she wants and how much he/she is willing to pay c) helps buyer compare houses that may be very different - in features and price. Potential benefits: happier and more financially stable buyers, decreased risk for mortgage provider, overall healthier real estate and mortgage system. (a few demo frames http://web.mit.edu/agchen/Public/CFB/House_base/bin-debug/testHouse.swf and http://web.mit.edu/agchen/Public/CFB/UtilityCalculator/bin-debug/UtilityCalculator.swf )


A New Type of Customer Network Visualization

Erik Feng, Dawei Shen, Kwan Hong Lee and Marko Popovic
(erikfeng[AT]MIT.EDU, dawei[AT]media.mit.edu, kwan[AT]media.mit.edu, marko__[AT]mit.edu)

Presentation http://web.media.mit.edu/~marko/ALL/FengShenLeePopovic.ppt
(if you can not access this link please click on title)

Physics based self-arranging dynamic graph used to visually identify groups of consumers based on their spending pattern. A graph is a network consisting of points or nodes (representing customers) connected by links (representing connections between the customers). Links act like nonlinear springs that stretch and shrink. System is scalable and may be used equally well with tens of nodes or millions of nodes. This physics based visual approach can be applied to numerous applications. (demo frame http://mit.edu/erikfeng/www/Restaurant.swf )


A New Type of Location Based Tracking System for Customer Spending

Ryan Macdowell and Marko Popovic
(rmacdo[AT]MIT.EDU,marko__[AT]mit.edu)

Presentation http://web.media.mit.edu/~marko/ALL/MacDowellPopovic.ppt
(if you can not access this link please click on title)

This web based application can be used by both the bank and customers as a financial assistant to better understand how, where, and when people spend their money. It could also be used to monitor, warn on preset goals and influence the spending habits. This animation utility automatically loads a set of transaction data into Google Maps and displays each transaction based on its location, size, type and the order in which the transactions occurred. System may be accessible by mobile devices such as the iPhone. System may improve fraud detection. Additionally, bank may view multiple users at once, compare them and show aggregate spending. (demo frame http://18.233.0.5/ccd/display.htm )


A New Type of Shopping Planner and Marketing Service

Andrew Okoh, Kwan Hong Lee and Marko Popovic
(manova[AT]MIT.EDU, kwan[AT]media.mit.edu, marko__[AT]mit.edu)

Presentation http://web.media.mit.edu/~marko/ALL/OkohLeePopovic.ppt
(if you can not access this link please click on title)

Shopping assistant application helps to optimize shopping experience based on parameters like cost, and proximity to store. Spending and savings is tracked in real time by lucid and interactive representations. Budget goals are determined for spending categories and alerts deployed when budget constraints are broken within specific categories. Application can provide a map platform for viewing stores with desired items, generate prioritized list of deals with highest priority given to cheaper items and alerts user if a store in close proximity is carrying out a sale on a specific item stored in a "wish list." In this way bank could provide additional service to consumers as well as merchants.


Lifestyle Visualization

Catherine Havasi, Ryan Macdowell, Rob Speer, Marko Popovic and Deb Roy
( havasi[AT]gmail.com, rmacdo[AT]MIT.EDU, marko__[AT]mit.edu, dkroy[AT]media.mit.edu )

Presentation http://web.media.mit.edu/~marko/ALL/HavasiMacDowellSpeerPopovicRoy.ppt
(if you can not access this link please click on title)

The Lifestyle Visualization project seeks to visually represent bank customers' spending habits as a function of time. By using common sense enriched data mining techniques, typical financial behaviors may be obtained from a large poll of customer data. Subsequently, each customer's spending trajectory can be represented in the behavioral cluster space in a visually appealing manner. The visual space is modeled after merchandise types, customer purchases and a common sense knowledge database. As a result, customer's motion is illustrated in the visual space as customer's spending habits change over time.


SmileMeter: A Smile Counter to achieve Customer Delight

Kyunghee Kim
( khkim[AT]MIT.EDU )

Presentation http://web.media.mit.edu/~marko/ALL/Kim.ppt
(if you can not access this link please click on title)

SmileMeter? is an intelligent vision system that provides information about nonverbal communication. SmileMeter? can be used in many areas in which social interactions happen such as banks, convenience stores, hospitals, schools, etc. The system is comprised of two main parts: 1. A Customer Affect Tagging System provides bank tellers with quantified nonverbal communication information such as facial expressions, head postures, eye gazes from the service encounters with their customers. 2. A Smile Counting Technology provides Bank of America with an innovative reward program in which the amount of awarded funds is proportional to the number of smiles from customers' facial expressions.


Mobile Banking Application

Steven Hong, Kwan Hong Lee and Marko Popovic
( hongs00[AT]MIT.EDU, kwan[AT]media.mit.edu, marko__[AT]mit.edu )

Presentation http://web.media.mit.edu/~marko/ALL/HongLeePopovic.ppt
(if you can not access this link please click on title)

Currently, mobile banking applications for the iPhone allow the user to view account balances, make transfers, and pay bills. The goal of this project is to expand ways to efficiently manage data on a small, multi-touch screen. This application provides a home menu allowing users to easily switch between options. Other features include graphic display of user account trends over time, placards providing more specific description of each transaction, and view transitions using iPhone's picker tool.


Customer Account Interface

Irum Amjad, Oscar Moll, and Marko B Popovic
( irumak[AT]mit.edu, rimoll[AT]MIT.EDU and marko__[AT]mit.edu )

Presentation http://web.media.mit.edu/~marko/ALL/AmjadMollPopovic.ppt
(if you can not access this link please click on title)

The goal of this project is to efficiently extract, classify, summarize and visualize customer spending over periods of time to benefit account holders and bank employees. A visual timeline (similar to RealPlayer?) displays spending for several categories (credit card payments, mortgage, insurance, student loan, food, entertainment, etc) as function of time. A variety of images (airplane, hotel, auto, etc) whose size is proportional to the amount spent reflects expenditures within each category (such as travel). This interface can be used to identify modes of account holder's pathological spending behaviors from current and historical data time patterns and resource allocations. Identifying this data can assist in predicting future spending patterns as well as preventing delinquency.


Banner Media Lab Home Page May 2008:
cfb-ban06.jpg


Seeking Undergraduate Research Oportunity Students

Faculty Supervisor: Professor Deb Roy

Direct Supervisor: Dr. Marko Popović

The Center for Future Banking is seeking Summer UROP students and graduate students interested in visualization and data mining of financial data. The research will be part of the Center for Future Banking, a new, innovative research collaboration with Bank of America.

Creative students with strong analytic skills and coding experience are preferred. Those working with us will have the opportunity to learn to use various visualization tools and gain data-mining skills.

Prerequisites: Experience with graphical tools, user interfaces, 3-D modeling, prefuse, and processing preferred. Coding experience required.

Undegraduate Students

  • Ryan Macdowell (rmacdo@mit.edu) SUMMER
  • Mary Hong (maryh10@mit.edu) SUMMER
  • Erik Feng (erikfeng@mit.edu) SUMMER
  • Steven Hong (hongs00@mit.edu) SUMMER
  • Leslie Gee (lgee4@mail.rochester.edu) SUMMER
  • Andrew Okoh (manova@mit.edu) SUMMER
  • Albert Chang (alchang@mit.edu)
  • Ana Chen (agchen@mit.edu) SUMMER
  • Oscar Ricardo Moll Thomae (rimoll@mit.edu) SUMMER
  • Iram Amjad(irumak@mit.edu) SUMMER

Graduate Students

  • Kwan Hong Lee (kwan@media.mit.edu)
  • Dawei Shen (dawei@media.mit.edu)
  • Kyunghee Kim (khkim@MIT.EDU)
  • Catherine Havasi (havasi@gmail.com)

Projects Manager

  • Marko B Popovic PhD (marko__@mit.edu)