Rich Holowczak's Research Interests

Financial Information Systems and Markets

My most recent work has been in the area of financial information systems and markets. With PhD student Norman Johnson (now at U. of Houston) we are investigating the differences in negotiation characteristics when different communications media (telephone, IM and quotation systems) are employed. With PhD student Ilan Levine, we are investigating applying market structures to the group decision support environment.

As Director of the Subotnick Financial Services Center, I am also looking into the how academic trading floors can be integrated into various curricula. Some publications from this research include:

A number of articles and white papers also metion our work at the Wasserman Trading Floor / Subotnick Center:

Data Warehousing and Data Mining in Environmental and Financial Digital Libraries

Data warehouses in digital libraries must contend with integrating and aggregating large amounts of multimedia data and should make this summarized information available in an efficient fashion. This research investigates a number of interesting research issues related to how multimedia data can be aggregated, summarized and indexed to allow for effective retrieval. Once data has been aggregated, data mining tools can then be applied to extract meaningful relationships and patterns.

A current project being developed at the Subotnick Center is the MarketVCR, a combination of a large multimedia data warehouse containing tick-by-tick quote and trade market data, video, audio and news articles, and a web application capable of tapping into the warehouse. Ultimately, the application will allow an instructor or student, in any location, at any point in the future, to “rewind” the market back to the time of the event and “play” the market with synchronized video and news releases just as it happened earlier in the day, week or year.

Extractors for Digital Library Objects

As the content in Digital Libraries grows, the ability to index and search becomes hampered because manual inspection of library objects (images, books, film, audio tapes etc.) becomes infeasible. My dissertation developed a collection of intelligent and domain specific Extractors that can extract concepts from digital library objects and make these concepts a part of the metadata used for indexing and querying.

One application I worked on uses Information Extraction techniques, a form of Natural Language Processing (NLP), that selectively identifies facts from free form text using semantic and syntactic patterns.

One project was done in conjunction with the Law Library of Congress' Global Legal Information Network (GLIN). This large collection of international laws provides an ideal "proving grounds" for a set of legal information extractors we are developing.

Another application of these techniques is in the area of financial reporting. We have done some preliminary work extracting various data from quarterly and annual reports (10-Q and 10-K reports) posted to the EDGAR system. Of particular interest is extracting the corporate subsidiary structure. some initial work in this area is reported here:

Digital Libraries and Electronic Commerce

Digital Libraries present unique research objectives that go beyond simply "digitizing traditional libraries." The vastly increased storage and network bandwidth requirements and broad range of potential users present problems in database design, network design, user interface design, distributed systems, information retrieval and indexing (to name but a few). Additional research issues can be found in our paper:

Current/Past Students

Here is a list of some of the student projects I have supervised:
Back to Rich's Home Page

File: research.html Last Update: March 7, 2008 © 1996-2008 R. Holowczak