Search

In this work we design a narrative understanding tool Text2ALM. This tool uses an action language ALM to perform inferences on complex interactions of events described in narratives. The methodology used to implement the Text2ALM system was originally outlined by Lierler, Inclezan, and Gelfond (2017) via a manual process of converting a narrative to an ALM model. It relies on a conglomeration of resources and techniques from two distinct fields of artificial intelligence, namely, natural language processing and knowledge representation and reasoning. The effectiveness of system Text2ALM is measured by its ability to correctly answer questions from the bAbI tasks published by Facebook Research in 2015. This tool matched or exceeded the performance of state-of-the-art machine learning methods in six of the seven tested tasks. We also illustrate that the Text2ALM approach generalizes to a broader spectrum of narratives., Comment: In Proceedings ICLP 2019, arXiv:1909.07646

This study presents new evidence on the relationship between high school inputs measured at the time male respondents attended high school and the earnings of these same individuals when they were in their mid-thirties. To accomplish this task, newly coded data on the characteristics of Wisconsin high schools in 1954-57 were matched to the Wisconsin Longitudinal Survey. Estimates show a significant relationship between the characteristics of teachers and the earnings of their students 17 years after graduation. A 1% increase in the average teacher salary in a district increases the earnings of students by 0.33%. The magnitude of this effect is larger than estimates reported in previous research, and many times larger than the impact of increasing parents' income by a comparable amount. (Contains 30 references.) (Author/(SLD)

This paper reports preliminary findings of a 3-year study examining the ability of qualitative and quantitative measures of the home environment variables (HEV's) to contribute to the prediction of third- through fifth-grade students' achievement and growth scores in reading and mathematics. Data were obtained for 198 parent/student pairs from four rural and small urban Wisconsin communities. Three sets each of achievement and growth scores were regressed on selected variables derived from parents' responses to a half-hour telephone interview. Each regression equation also contained a measure of the child's academic aptitude as the predominant predictor. Several HEV's provided significant and durable additional predictive capacity to that provided by the aptitude index. These included (1) hours of maternal employment (negatively associated with all students' and boys' achievement in both reading and math); (2) family socioeconomic status (positively associated with all students' and boys' reading growth and all students' reading achievement); and (3) parents' feeling that the school's greatest strength was the quality of its academic programs (positively associated with girls' reading and math achievement and math growth). Also examined was the indirect link between home environment and academic attainment through HEV associations with self-concept variables related to achievement. In this case, no substantial link was identified. (RH)