Hierarchical text classification using Relative Inverse Document Frequency

Automatic hierarchical text classification has been a challenging and in-needed task with an increasing of hierarchical taxonomy from the booming of knowledge organization. The hierarchical structure identifies the relationships of dependence between different categories in which can be overlapped of generalized and specific concepts within the tree. This paper presents the use of frequency of the occurring term in related categories among the hierarchical tree to help in document classification. The four extended term weighting of Relative Inverse Document Frequency (IDFr) including its located category, its parent category, its sibling categories and its child categories are exploited to generate a classifier model using centroid-based technique. From the experiment on hierarchical text classification of Thai documents, the IDFr achieved the best accuracy and F-measure as 53.65% and 50.80% in Top-n features set from family-based evaluation in which are higher than TF-IDF for 2.35% and 1.15% in the same settings, respectively.

The use of relative inverse thermal admittance for the characterization and optimization of fin-wall assemblies

The concept of relative inverse thermal admittance applied to the convective fin-wall assembly optimization of longitudinal rectangular fins under 2-D heat conduction is presented in this work. Since heat transfer at the fin tip is taken into account, it is not always possible to optimize the above cited geometry. This is relevant in optimization processes and because of this has been displayed in several graphs. Here, different values for convective conditions at the fin and wall surfaces are used and the influence of the hw/hf ratio in optimum geometry is determined. The fin effectiveness is used as the fundamental parameter to prove that the fin is fulfilling the objective of increasing heat dissipation. Once the optimum thickness has been obtained, the Biot number is easily calculated and the fin effectiveness for an isolated fin and the fin-wall assembly can be determined graphically. The optimization process is carried out through a set of universal graphs in which the range of parameters covers most of the practical cases a designer will find. The concept of relative inverse thermal admittance is applied in a general form and emerges as an easy used tool for optimizing fin-wall assemblies.

Design and Optimization of Composite Rectangular Fins Using the Relative Inverse Thermal Admittance

The concept of relative inverse admittance applied to composite fins optimization in the case of longitudinal rectangular fins under 2D heat conduction is presented in this work. Here, different values for convective conditions at the fin and composite layer surfaces are used and the influence of the kc/kf ratio and composite thickness in optimum geometry is determined. The optimization process is carried out through universal graphs in which the range of parameters covers most of the practical cases a designer will find. Relative inverse admittance is applied in a general form and emerges as an easily used tool for optimizing composite fins.

Optimization of Longitudinal Rectangular Fins With Asymmetrical Boundary Conditions

Longitudinal rectangular fins with asymmetrical boundary conditions and different temperatures at each base were been studied and optimized using relative inverse thermal admittance as performance coefficient. A Neutral Line was defined and determined for different temperature ratios at the bases. Using the network method for the numerical simulation, the optimization problem was solved in a classical way: starting with the fin volume, heat transfer coefficient and thermal conductivity, the optimal geometry or aspect ratio (length/thickness ratio) for dissipating maximum heat from the fin surface was determined. The results are illustrated through tables and universal graphics made with optimal points, which, are independent from the temperature ratio at the bases. Besides the optimal thickness the graphics also provide, the fin effectiveness and the transversal Biot number. Finally, the location of the Neutral Line is represented for the optimal geometry and temperature ratios. To illustrate the optimization process practical examples are shown.

Universal Curves for Optimizing Longitudinal Rectangular Fins Radiating to Free Space

The relative inverse admittance is used to obtain the geometry of the optimum longitudinal rectangular fin of a set of fins of the same volume that radiate to free space. With this information a set of new universal curves that permits direct optimization of 1-D and 2-D practical fins. To ensure a dissipation effect, lines of constant effectiveness are introduced in the optimization graph. Examples are shown.

An Exploration of the Nature and Rationale of a Systemic Model

A somewhat abstruse model recently formulated by Alonso is introduced and analyzed. The nature and certain structural properties of the model are examined and its general form is identified. The salient feature of the model is its simultaneous and consistent determination of the totals of the endogenous variables and of their constituent parts. The model can be viewed as an integration of substantive (producing–causal) submodels with a structural–formal (distributive) submodel. The integration is based on the dual roles of two endogenous variables both as normalizing factors and as explanatory factors. The last role is justified by the factors' representation of relative (inverse) mean cost of interaction. A statistical-independence postulation is also identified as an underlying statement of the model.

Completely simple and inverse semigroups

The purpose of this paper is to investigate the structure of certain types of semigroups. Rees(6),(7) has determined the structure of a completely simple semigroup, and has shown that such a system may be realized as a type of matrix semigroup. Clifford (2) and Schwarz (8) have found conditions, namely, the existence of minimal left and minimal right ideals, under which a simple semigroup is completely simple, and have made a more detailed study of such semigroups. Preston (4), (5) has studied inverse semigroups, in which each non-zero element has a unique relative inverse, and has also considered inverse semigroups which contain minimal right or left ideals. In the present paper we obtain a set of conditions on a simple semigroup, each of which is equivalent to the semigroup being both completely simple and inverse. Section 2 defines the terms used and gives a brief resume of the main results which have already been proved. Section 3 is devoted to our present considerations.