Modeling Expertise through Biographically Plausible Datasets
Colin Allen, Cognitive Science Program
Digitized materials from IU’s Hathi Trust Research Center will be used by the InPhO group to build cognitively rich models of expert knowledge. We aim to model “biographically plausible” datasets, i.e., sets of books that a person could plausibly read in a lifetime. We hypothesise that such expertise-based models of the collection will serve practical and scholarly purposes better than a generic models of the whole. For this pilot project we will test this hypothesis using against several modeling techniques with a goal of extending those techniques. We aim to showcase the effectiveness of our methods on a scientifically and historically important set of documents, specifically Charles Darwin’s library.
Serial Reproduction of Narratives
Fritz Breithaupt, Department of Germanic Studies
The project is to conduct foundational research in the methodology of serial reproduction of narratives (telephone games with narrative sequences of 3-6 sentences). The approach has been used occasionally since the pioneering study by Frederic Bartlett (1932). However, there are few studies that have tested the approach with any systematic depth. Whereas basic story events (hero kills dragon, girl steals apple…) tend to be preserved well, most other elements are fair game to be dropped, changed, or invented. The changes reflect impacts of preference, moral choice, memory, optimal narratives, cultural expectations, etc. Our work will focus on preservation of events.
Dual Mechanism of Action Anti-Tuberculosis Compounds Through Artemisinin Conjugation
Silas Cook, Department of Chemistry
Tuberculosis (TB) is a major global health threat. In this proposal, both known and experimental anti-TB drugs will be covalently linked, or conjugated, to the antimalarial drug artemisinin. In malaria parasites, artemisinin reacts with iron to initiate a radical cascade that rapidly kills the cell. By covalently linking artemisinin to drugs that target Mycobacterium tuberculosis (Mtb), the cell-killing action of artemisinin can be brought to bear on Mtb. This “artemisinin conjugation” approach to generate anti-TB agents will provide a new platform for the rapid enhancement of anti-TB potency for compounds with known anti-TB activity.
Preclinical Studies of Cannabis Abuse
Ken Mackie, Department of Psychological & Brain Sciences
This is an FRSP proposal to support pilot studies using a preclinical model of Cannabis abuse. Its primary goal is to obtain preliminary data for several NIH grant submissions. Heavy early adolescent Cannabis use increases the risk of later psychotic disorders. Our hypothesis is that tetrahydrocannabinol (THC) from Cannabis activates prefrontal cortical (PFC) microglial cells during a particularly vulnerable period of the PFC's development (adolescence). The activated microglia then inappropriately prune synapses in the maturing PFC (a time of active synaptic remodeling). This alters the number of synaptic connections onto PFC neurons, leading to an imbalance of excitatory and inhibitory synaptic transmission onto these neurons.
Construction of a Linguistic Corpus and Ontology for Comparative Analysis of Networks
Armando Razo, Department of Political Science
Markus Dickinson, Department of Linguistics
This project establishes the foundations for an innovative approach to enable large-scale comparative social science research via the use of web-based ontologies. We plan to systematize body of knowledge of networks in comparative politics and international development by developing novel computational methods to extract both theoretical and applied meaning from unstructured textual sources. In particular, we plan to produce two deliverables: (1) an original corpus of theoretical and applied research from an existing database of scholarly work on political networks and international development; and, with added natural language processing of that corpus, (2) a pilot web-based ontology to be scaled up with subsequent external funding.
A Sensitive Search for Dark Energy Through Chameleon Scalar Fields using Neutron Interferometry
William Michael Snow, Department of Physics
The physical origin of the dark energy which is postulated to cause the accelerated expansion rate of the universe is one of the major open questions of cosmology. We describe a measurement using neutron interferometry which can place the most stringent upper bound on "chameleon fields" proposed as a solution to the problem of the origin of dark energy of the universe.