Garett O. Dworman

On automated discovery of models using genetic programming: bargaining in a three-agent coalitions game

The creation of mathematical, as well as qualitative (or rule-based), models is difficult, time-consuming, and expensive. Recent developments in evolutionary computation hold out the prospect that, for many problems of practical import, machine learning techniques can be used to discover useful models automatically. The prospects are particularly bright, we believe, for such automated discoveries in the context of game theory. This paper reports on a series of successful experiments in which we used a genetic programming regime to discover high-quality negotiation policies. The game-theoretic context in which we conducted these experiments-- a three-player coalitions game with sidepayments--is considerably more complex and subtle than any reported in the previous literature on machine learning applied to game theory.

Helping users to use help: improving interaction with help systems

The content management, IA, and documentation communities have written extensively about the structure and content of help systems, as well as on help delivery mechanisms. But even if content and IA are perfectly suited to users’ needs and tastes, a help system may still fail to engage users, because they don’t interact with the help system in the first place. This workshop focuses on effective integration of help systems into users’ environments. It will help us understand how users access help systems, so we can improve the initial interaction between users and help systems.

On automated discovery of models using genetic programming in game-theoretic contexts

The creation of mathematical, as well as qualitative (or rule-based), models is difficult, time-consuming, and expensive. Recent developments in evolutionary computation hold out the prospect that, for many problems of practical import, machine learning techniques can be used to discover useful models automatically. These prospects are particularly bright, we believe, for such automated discoveries in the context of game theory. This paper reports on a series of successful experiments in which we used a genetic programming regime to discover high-quality negotiation policies. The game-theoretic context in which we conducted these experiments-a three-player coalitions game with sidepayments-is considerably more complex and subtle than any reported in the literature on machine learning applied to game theory.<<ETX>>

MOTC: an interactive aid for multidimensional hypothesis generation

The paper reports on conceptual development in the areas of database mining and knowledge discovery in databases (KDD). Our efforts have also led to a prototype implementation, called MOTC, for exploring hypothesis space in large and complex data sets. Our KDD conceptual development rests on two main principles. First, we use the crosstab representation for working with qualitative data. This is by now standard in on-line analytical processing (OLAP) applications, and we reaffirm it with additional reasons. Second, and innovatively, we use prediction analysis as a measure of goodness for hypotheses. Prediction analysis is an established statistical technique for analysis of associations among qualitative variables. It generalizes and subsumes a large number of other such measures of association, depending on specific assumptions the user is willing to make. As such, it provides a very useful framework for exploring hypothesis space in a KDD context. The paper illustrates these points with an extensive discussion of MOTC.

On pattern-directed search of archives and collections

This article begins by presenting and discussing the distinction between record-oriented and pattern-oriented search. Examples of record-oriented (or item-oriented) questions include: “What (or how many, etc.) glass items made prior to 100 A.D. do we have in our collection?” and “How many paintings featuring dogs do we have that were painted during the 19th century, and who painted them?” Standard database systems are well suited to answering such questions, based on the data in, for example, a collections management system. Examples of pattern-oriented questions include: “How does the (apparent) production of glass objects vary over time between 400 B.C. and 100 A.D.?” and “What other animals are present in paintings with dogs (painted during the 19th century and in our collection)?” Standard database systems are not well suited to answering these sorts of questions (and pattern-oriented questions in general), even though the basic data is properly stored in them. To answer pattern-oriented questions it is the accepted solution to transform the underlying (relational) data to what is called the data cube or cross tabulation form (there are other forms as well). We discuss how this can be done for non-numeric data, such as are found widely in museum collections and archives. Further we discuss and demonstrate two distinct, but related, approaches to exploring for patterns in such cross tabulated museum data. The two approaches have been implemented as the prototype systems Homer and MOTC. We conclude by discussing initial experimental evidence indicating that these approaches are indeed effective in helping people find answers to their pattern-oriented questions of museum and archive collections. Two Kinds of Questions One’s purpose, when approaching an archive or museum collection for information, might be characterized as seeking an answer to one or more questions. Thus, if an information system is to be helpful in answering one’s questions of archives and collections, it would seem that categorizing the questions to be asked can only be helpful in designing an information system to assist in answering them. What kinds of questions are there that are pertinent to archives and museum collections? This is a large and difficult issue, and we do not expect to resolve it here. Our aim in this article is more modest: we wish to distinguish two kinds of questions, and to explore their relevance to museum and archive informatics. We devote the remainder of the present section to making and exploring our basic distinction. The sections that follow explore the distinction in the context of a particular information system, the Core of Discovery system, installed at The Historic New Orleans Collection. The distinction we wish to make here, and to exploit in designing museum and archive information systems, is deeply embedded in folklore and ordinary language. “You cannot see the wood for the trees” is perhaps the earliest recorded embodiment of the distinction in English. [The quotation is from John Heywood’s Proverbs, itself the earliest published (1546) collection of English folk sayings.] Proverbially, there is a distinction to be made between seeing (or asking about) the trees and seeing (or asking about) the forest. But how can we characterize the distinction and what can we do to provide computerized support for these two kinds of questions? One question at a time. First a characterization of the distinction. The distinction is best seen through a series of examples. Let us compare some tree questions with some forest questions. Here are some questions about trees in a forest.

The internal consultancy model for strategic UXD relevance

Experts in the field of HCI have spoken at length about how to increase the strategic influence of User Experience Design (UXD) teams in industry [3]. Some have offered courses in HCI management [1]. Others have presented recommendations on how to prove a return on investment for usability-related activities [2]. Nielsen [5] has described the usability maturity model, presenting implicit management challenges and structures at different phases.Few though have discussed the value and process for an embedded UXD Group functioning as an internal consultancy to different product teams within their organizations, and how this model can increase the strategic relevance of UXD in their companies. The Cisco UXD Group evolved through several funding and organizational models (central funding, client-funding, distributed teams), and now follows an internal consultancy model. This paper describes the experiences that led to this model and how it has helped increase the strategic influence of UXD within Cisco.

MOTC: an aid to multidimensional hypothesis generation

Reports on conceptual development in the areas of database mining and knowledge discovery in databases (KDD). Our efforts have also led to a prototype implementation, called MOTC, for exploring hypothesis spaces in large and complex data sets. Our KDD conceptual development rests on two main principles. First, we use the crosstab representation for working with qualitative data. This is by now standard practice in OLAP (online analytical processing) applications, and we reaffirm it with additional reasons. Second, and innovatively, we use prediction analysis as a measure of goodness for hypotheses. Prediction analysis is an established statistical technique for analysis of associations among qualitative variables. It generalizes and subsumes a large number of other such measures of association, depending upon the specific assumptions the user is willing to make. As such, it provides a very useful framework for exploring the hypothesis space in a KDD context. This paper illustrates these points with an extensive discussion of MOTC.

Special interest group for the CHI 2010 management community

This SIG will provide an opportunity for those interested in the relationship between management and HCI to explore this subject and the ongoing development of the Management Community at CHI conferences and beyond.

Comparing internal UXD business models

Experts in the field of HCI have spoken at length about how to increase the strategic influence of User Experience Design (UXD) teams in industry [3]. Some have offered courses in HCI management [1]. Others have presented recommendations on how to prove a return on investment for usability-related activities [2]. This SIG is an extension of the CHI experience report.The Internal Consultancy Model for Strategic UXD Relevance,. [5] and explores four common UXD organizational models. In this SIG, we will develop a SWOT analysis (analyzing Strengths, Weaknesses, Opportunities, and Threats) of each model. The SIG will facilitate a systematic exploration by attendees whose organizations follow, or are considering, one or more of these models. It will result in a broader understanding for managers of UXD teams on how they can optimally structure their internal UXD functions, given their unique corporate environments and cultures.

MOTC with Examples: An Interactive Aid for Multidimensional Hypothesis Generation ¤

This paper reports on conceptual development in the areas of database mining, and knowledge discovery in databases (KDD). Our efforts have also led to a prototype implementation, called MOTC, for exploring hypothesis space in large and complex data sets. Our KDD conceptual development rests on two main principles. First, we use the crosstab representation for working with qualitative data. This is by now standard practice in OLAP (on-line analytical processing) applications and we reaffirm it with additional reasons. Second, and innovatively, we use Prediction Analysis as a measure of goodness for hypotheses. Prediction Analysis is an established statistical technique for analysis of associations among qualitative variables. It generalizes and subsumes a large number of other such measures of association, depending upon specific assumptions the user is willing to make. As such, it provides a very useful framework for exploring hypothesis space in a KDD context. The paper illustrates these points with an extensive discussion of MOTC.