This paper poses the challenge of developing and evaluating integrated systems for computational scientific discovery. We note some distinguishing characteristics of discovery tasks, examine eight component abilities, review previous successes at partial integration, and consider hurdles the AI research community must leap to transform the vision for integrated discovery into reality. In closing, we discuss promising scientific domains in which to test such computational artifacts.
The importance and applications of nanoscale magnetic storage devices has been of current interest owing to the extensive research interest in magnetic nanocaps. Continuing their research trend, we have attempted to syn-thesize and characterize an important isolated magnetic nanocaps system constituting Co and its oxide. CoO/Co thin films were deposited onto self-assembled arrays of polystyrene (PS) nanospheres (similar to 600 nm diameter) under ultra-high vacuum environment using electron beam evaporation technique. The magnetic and structural properties of these nanostructures were then compared with those of simultaneously deposited films on bare Si substrate (referred to as reference film). The studied film and the reference film were grown in polycrystalline manner as observed from X-ray diffraction measurements while their roughnesses as observed from X-ray reflectivity were quite different. X-ray reflectivity showed that for the reference films well defined Kiessig os-cillations appeared suggesting low roughness in the deposited films, while the films on PS followed the curvature of underlying nanospheres and thus have very high roughness resulting in the disappearance of Kiessig oscil-lations. Magnetic measurements exhibited a drastically high coercivity when the substrate was changed from flat (Si) to curved one (PS). As the film thickness was increased, the coercivity first showed a slight decrement (4.92 kA/m) and beyond 40 nm film thickness, it showed some enhancement (14.2 kA/m). The exchange bias mea-surements also showed interesting results with variation in film thickness. Co 10 nm film deposited on PS showed negative exchange bias of-127 kA/m which decreased to-26 kA/m in 100 nm film. The overall results were explained by correlating the magnetic and microstructural properties of the thin films as a function of thickness.
This paper presents a novel approach to the acquisition of language models from corpora. The framework builds on Cobweb, an early system for constructing taxonomic hierarchies of probabilistic concepts that used a tabular, attribute-value encoding of training cases and concepts, making it unsuitable for sequential input like language. In response, we explore three new extensions to Cobweb -- the Word, Leaf, and Path variants. These systems encode each training case as an anchor word and surrounding context words, and they store probabilistic descriptions of concepts as distributions over anchor and context information. As in the original Cobweb, a performance element sorts a new instance downward through the hierarchy and uses the final node to predict missing features. Learning is interleaved with performance, updating concept probabilities and hierarchy structure as classification occurs. Thus, the new approaches process training cases in an incremental, online manner that it very different from most methods for statistical language learning. We examine how well the three variants place synonyms together and keep homonyms apart, their ability to recall synonyms as a function of training set size, and their training efficiency. Finally, we discuss related work on incremental learning and directions for further research.
The Polyol chemical synthesis method being a simple and economic method for preparation of nanoparticles (NP) was used to prepare Iron Pyrite (FeS2) NP with surfactant octylamine and precursor Thiourea. The crystalline quality and stoichiometry of these NPs were confirmed by X-ray diffraction and Raman Spectroscopy, which gave an average crystallite size of ∼30 nm. The sharp well-defined nature of the peaks indicated good crystallinity of the samples. The Rietveld refinement of the diffraction data revealed the lattice parameters, hkl values, phase, Profile Factor etc. of FeS2 NP. The elemental composition was determined using Energy-Dispersive X-ray Spectroscopy, where characteristic emission peaks of Fe and S elements without other chemical impurities were observed. Further, Raman Spectroscopy provided information on chemical bonding and symmetry of molecules in these surfactant coated FeS2 NP. The strain introduced in these NP during synthesis is also discussed.
