Supervised clustering for automated document classification and prioritization: a case study using toxicological abstracts

Machine learning and natural language processing algorithms are currently widely used to retrieve relevant documents in a variety of contexts, including literature review and systematic review. Supervised machine learning algorithms perform well in terms of retrieval metrics such as recall and precision, but require the use of a sizeable training dataset, which is typically expensive to develop. Unsupervised machine learning algorithms do not require a training dataset and may perform well in terms of recall, but are typically lower in precision, and do not offer a transparent means for decision-makers to justify selection choices. In this paper, we illustrate the use of a hybrid document classification method based on semi-supervised learning that we refer to as “supervised clustering.” We show that supervised clustering combines the ease of use of unsupervised algorithms with the retrieval efficiency and transparency of supervised algorithms. We demonstrate through simulations the high performance and unbiased predictions of supervised clustering when provided even with only minimal training data. We further propose the use of ensemble learning as a means to maximize retrieval efficiency and to prioritize the review of those documents that are not eliminated by the supervised clustering algorithm.     

Methodology for consequence analysis of LNG releases at deepwater port facilities

A methodology to perform consequence analysis associated with liquefied natural gas (LNG) for a deepwater port (DWP) facility has been presented. Analytical models used to describe the unconfined spill dynamics of LNG are discussed. How to determine the thermal hazard associated with a potential pool fire involving spilled LNG is also presented. Another hazard associated with potential releases of LNG is the dispersion of the LNG vapor. An approach using computational fluid dynamics tools (CFD) is presented. The CFD dispersion methodology is benchmarked against available test data. Using the proposed analysis approach provides estimates of hazard zones associated with newly proposed LNG deepwater ports and their potential impact to the public.     

Further Results on Permit Markets with Market Power and Cheating

This paper examines a market for pollution permits in which one firm has market power and one or more firms is noncompliant. I show that the firm with market power may choose to hold more permits than it needs, effectively retiring permits from the market. I also show that some noncompliance may be socially desirable because it can mitigate the distortion caused by market power. Similarly, some degree of market power may be socially desirable because it can, in turn, mitigate the distortion caused by noncompliance.     

Effect of application of ammonium thiosulphate on production and emission of methane in a tropical rice soil

Agricultural sources of atmospheric methane include flooded rice (Oryza sativa L.) paddies. However, certain soil nutrient management and cultural practices offer opportunities to reduce methane emissions. The effect of application of ammonium thiosulphate, a potential source of nitrogen and sulphur and also an inhibitor of nitrification and urease on methane production and emission from flooded alluvial (Typic Haplaquept) rice soil in India, was examined. Methane production and emission from control and urea-amended soil samples were almost identical. Application of ammonium thiosulphate to laboratory-incubated flooded soil (30 and 60 μg N g⁻¹ soil) and flooded rice fields (45.6 and 60 kg N ha⁻¹) effected a distinct inhibition of methane production and emission. Ammonium thiosulphate stimulated the population of sulphate-reducing bacteria (SRB) to a greater extent at 60 μg N g⁻¹ soil than at 30 μg N g⁻¹ soil. In ammonium thiosulphate-applied rice field plots, mean methane efflux decreased by about 38 and 60% at 45.6 and 60 kg N ha⁻¹, respectively, over that of control. Inhibition of methane production by ammonium thiosulphate is, at least in part, due to the stimulation of SRB. Results suggest the mitigation potential of ammonium thiosulphate on methane emission from flooded rice paddies.     

Molecular level toxicological evaluation of a dental composite implanted in albino rats

Dental composite (DC) resins are synthetic resins which are used as restorative material or adhesives in dentistry. Monomers and other components released into the oral environment from the DC material even after polymerization might affect the surrounding tissues and even the whole body gradually. Biocompatibility of DC has to be stringently evaluated since the fillings are in close contact with mucosa, tooth, and pulp. Common difficulties with various current dental materials include allergies, chemical leakage from the material, and pulpal irritation which are due to the byproducts of chemical reactions during different stages of material hardening. Dental Products Laboratory of Sree Chitra Tirunal Institute for Medical Sciences and Technology developed a new BIS-GMA-based DC. In this study the oxidative stress induced by this DC material in vivo was tested for a period of 1, 4, 8, 26, or 52 weeks in albino rats. Dental materials were implanted intramuscularly into the gluteus muscle and oxidative biomarkers such as lipid peroxidation, 8-oxo-deoxyguanosine (8-OHdG) concentration, and total glutathione content analyzed. Data showed no significant changes in antioxidant defense system and 8-OHdG levels in nuclear DNA of the tissue. Serum biochemical results confirmed that animals were in normal functioning physiological state.     

Comparative cytotoxicity of nanoparticles and ions to Escherichia coli in binary mixtures

The objective of this study was to understand toxicity of mixture of nanoparticles (NPs) (ZnO and TiO₂) and their ions to Escherichia coli. Results indicated the decrease in percentage growth of E. coli with the increase in concentration of NPs both in single and mixture setups. Even a small concentration of 1 mg/L was observed to be significantly toxic to E. coli in binary mixture setup (exposure concentration: 1 mg/L ZnO and 1 mg/L TiO₂; 21.15% decrease in plate count concentration with respect to control). Exposure of E. coli to mixture of NPs at 1000 mg/L (i.e., 1000 mg/L ZnO and 1000 mg/L TiO₂) resulted in 99.63% decrease in plate count concentration with respect to control. Toxic effects of ions to E. coli were found to be lesser than their corresponding NPs. The percentage growth reduction was found to be 36% for binary mixture of zinc and titanium ions at the highest concentration (i.e., 803.0 mg/L Zn and 593.3 mg/L Ti where ion concentrations are equal to the Zn ions present in 1000 mg/L ZnO NP solution and Ti^+ 4 ions present in 1000 mg/L TiO₂ NP solution). Nature of mixture toxicity of the two NPs to E. coli was found to be antagonistic. The alkaline phosphatase (Alp) assay indicated that the maximum damage was observed when E. coli was exposed to 1000 mg/L of mixture of NPs. This study tries to fill the knowledge gap on information of toxicity of mixture of NPs to bacteria which has not been reported earlier.