Automating argument construction

Over the past five years the Artificial Intelligence Center at SRI has been developing a new technology to address the problem of automated information management within real- world contexts. The result of this work is a body of techniques for automated reasoning from evidence that we call evidential reasoning. The techniques are based upon the mathematics of belief functions developed by Dempster and Shafer and have been successfully applied to a variety of problems including computer vision, multisensor integration, and intelligence analysis.

We have developed both a formal basis and a framework for implementating automated reasoning systems based upon these techniques. Both the formal and practical approach can be divided into four parts: (1) specifying a set of distinct propositional spaces, (2) specifying the interrelationships among these spaces, (3) representing bodies of evidence as belief distributions, and (4) establishing paths of the bodies for evidence to move through these spaces by means of evidential operations, eventually converging on spaces where the target questions can be answered. These steps specify a means for arguing from multiple bodies of evidence toward a particular (probabilistic) conclusion. Argument construction is the process by which such evidential analyses are constructed and is the analogue of constructing proof trees in a logical context.

This technology features the ability to reason from uncertain, incomplete, and occasionally inaccurate information based upon seven evidential operations: fusion, discounting, translation, projection, summarization, interpretation, and gisting. These operation are theoretically sound but have intuitive appeal as well.

In implementing this formal approach, we have found that evidential arguments can be represented as graphs. To support the construction, modification, and interrogation of evidential arguments, we have developed Gister. Gister provides an interactive, menu-driven, graphical interface that allows these graphical structures to be easily manipulated.

Our goal is to provide effective automated aids to domain experts for argument construction. Gister represents our first attempt at such an aid.     

Sorokin as Lifelong Russian Intellectual: The Enactment of an Historically Rooted Sensibility

Prior to his 1922 emigration to Europe and thence to the United States, Pitirim Alexandrovich Sorokin had an exceptional intellectual and political career in Russia and the Soviet Union (Sorokin 1924, 1963a; Johnston 1995; Krotov 2005). Indeed, he was among the early founders of the science of sociology in his native land, where, according to a relatively recent bibliography (Sorokin 2000), he produced 162 Russian-language publications between the ages of 21 and 33. This listing includes not only book reviews and journal articles, but also substantial monographs and a two-volume theoretical treatise. While still a relatively young man, Sorokin had thus gained widespread recognition as a scholar of the first rank. He was also the initial chairperson (from 1919 to 1922) of a fledgling department of sociology at the University of Petrograd (St. Petersburg), an elected member of the national Constituent Assembly and an appointed staff member of the 1917 Provisional Government, the first democratic regime in Russia. This much would have sufficed for an entry in a sociological encyclopedia, and Sorokin’s political career has few parallels in the history of the field, other than the involvement of Emile Durkheim in French educational policy and the participation of Max Weber in creating the Weimar Republic in Germany. Nevertheless sociologists in the United States and most western historians of the field have not yet appreciated the full influence of the formative period, especially from 1905 to 1922. Lacking familiarity with Russian culture of that era and knowing little about the larger Russian socio-historical milieu, its intellectual discourse and collective memory, they have not been able to comprehend Sorokin’s outlook, behavior and professional output in the United States in relation to these earlier contextual factors. This is arguably a fundamental reason why many U.S. sociologists have tended to see Sorokin, especially since 1937, as a marginal figure and to regard his works largely as deviations from accepted social scientific practice. This paper will argue that a more adequate appreciation of Sorokin’s background and early adult life illumines both stylistic features of his works in America and also places into proper perspective several of his substantive foci that did not accord with contemporary “normal science” (Kuhn 1962). In short, despite his overall assimilation into American society and higher education, including his appointment at Harvard University and his election as president of the American Sociological Association, Sorokin should be understood in large measure as a life-long Russian intellectual. His was a Russian-born sensibility and consciousness—indeed a “Russian soul”—so deeply ingrained that it stamped his entire professional career in the United States, including his published researches, his popular sociology and his university teaching.     

Strength in Diversity? Group Heterogeneity in the Mobilization of Grassroots Organizations

Grassroots groups often find that they need to manage issues of diversity along racial/ethnic, class, gender, religious, and/or citizenship lines, among several other axes of difference. There are strong reasons to expect that diversity is a benefit to such groups, especially in terms of coalition breadth and the widely held expectation that internal heterogeneity helps organizations to resolve collective action dilemmas. Nonetheless, a long tradition within sociological research sets out the expectation that group heterogeneity thwarts the formation of collective identities and group solidarity, and therefore limits the potential for effective organizing. With a particular emphasis on groups active in local community‐based organizing since the 1970s, we review studies on a variety of features of group diversity and their effects on mobilizing processes. We find that although researchers remain somewhat divided on these critical issues, the preponderance of evidence suggests that managing diversity poses significant challenges for grassroots organizing efforts. While these challenges may not easily be overcome, we point to a variety of cases in which group differences offered a significant asset to social change efforts.     

Bidirectional Relations Between Authoritative Parenting and Adolescents’ Prosocial Behaviors

This study examined the bidirectional relations between authoritative parenting and adolescents’ prosocial behavior over a 1‐year time period. Data were taken from Time 2 and 3 of the Flourishing Families Project, and included reports from 319 two‐parent families with an adolescent child (M age of child at Time 2 = 12.34, SD = 1.06, 52% girls). Cross‐lag analyses supported bidirectional relations between parenting and prosocial behavior with particular emphasis on the role of the adolescents’ prosocial behavior on subsequent parenting. Results also varied as a function of the reporter. Discussion focuses on the implications for understanding the multifaceted nature of prosocial development in adolescence.     

Racial/Ethnic Patterns of Kindergarten School Enrollment in the United States

Enrollment into unequal schools at the start of formal education is an important mechanism for the reproduction of racial/ethnic educational inequalities. We examine whether there are racial/ethnic differences in school enrollment options at kindergarten, the start of schooling. We use nationally representative data from the Early Childhood Longitudinal Study‐Birth Cohort (ECLS‐B) to model whether parents seek information about their child's school before enrolling, whether parents move to a location so that a child can attend a certain school, or whether parents enroll their child in a school other than the assigned public school. Results indicate that enrollment patterns differ greatly across race/ethnicity. Whereas Black families are the most likely to seek information on a school's performance, White families are the most likely to use the elite option of choosing their residential location to access a particular school. These differences persist when controlling for socioeconomic status and sociogeographic location. Kindergarten enrollment patterns preserve the advantages of White families, perpetuating racial/ethnic disparities through multiple institutions and contributing to intergenerational processes of social stratification. Research should continue to examine specific educational consequences of housing inequities and residential segregation.     

BUSINESS COMBINATION DECISIONS: A U.S./CANADA STUDY*

The purpose of this paper is to clarify the motives behind business combination decisions. Twenty-one favorable characteristics associated with the motivations for business combinations were selected to serve as the basis for an empirical survey. Our analysis revealed that there are seven motivational variables of varying import and that the motivations underlying business combination decisions are similar for U.S. and Canadian executives.     

A Bayesian approach to non-parametric monotone function estimation

Summary.  The paper proposes two Bayesian approaches to non-parametric monotone function estimation. The first approach uses a hierarchical Bayes framework and a characterization of smooth monotone functions given by Ramsay that allows unconstrained estimation. The second approach uses a Bayesian regression spline model of Smith and Kohn with a mixture distribution of constrained normal distributions as the prior for the regression coefficients to ensure the monotonicity of the resulting function estimate. The small sample properties of the two function estimators across a range of functions are provided via simulation and compared with existing methods. Asymptotic results are also given that show that Bayesian methods provide consistent function estimators for a large class of smooth functions. An example is provided involving economic demand functions that illustrates the application of the constrained regression spline estimator in the context of a multiple-regression model where two functions are constrained to be monotone.