Irrelevant Ones Research Articles

Semantic and Pragmatic Features of Hadrami Jokes

Humour with its different forms is a universal phenomenon. The joke is among these forms. Although they are not famous for their sense of humour, Hadrami people have their own special cultural heritage of jokes. This study attempts to investigate the semantic and pragmatic features of the punchline of a selected number of Hadrami jokes depending on two parameters of the General Theory of Verbal Humour (GTVH) which are Script Opposition (SO) and Language (LA). The linguistic constructs of LA are adopted from Leech's (1981) associative meanings and Grice's (1975) conversational implicature theory. The sample has been collected from the Internet and Hadrami people. After excluding the irrelevant ones, 37 jokes has been studied within the theoretical framework of this study following the discourse analysis approach of linguistic qualitative researches. The findings show that the most frequent SO in Hadrami jokes are: expected/ unexpected, and stingy/ not stingy. Also, the semantic features exceed the pragmatic ones. Within the semantic features, the connotations exceed – in their turn- the social and the collocative meanings. The focus of the pragmatic features is violating the maxims mainly under the SO parameter stingy/ not stingy. These findings prove that the punchlines of Hadrami jokes are the focal point of the joke and they represent loaded language with a great diversity of semantic and pragmatic indications.

Assessment of the relevance of the student survey results in the educational process

Many educational institutions implement feedback mechanisms in their activities, this also occurs according to the requirements of educational standards, public accreditation organizations, international quality standards. But it is not always clear how to competently and effectively use the data obtained due to the unsuitability of such data in its original raw form. The study is aimed at developing a methodology for filtering student reviews about the studied disciplines from irrelevant ones in order to further use them in the management system of an educational institution and the educational process. The study used the methods of statistical data analysis and estimation of the parameters of a logistic regression model using STATISTICA 10.0 and Jamovi 1.8.1 programs. Spearman's and Kendall's correlation coefficients were used to assess the direction and strength of the link between quantitative indicators. Cramer's V test was used to assess the relationship between dichotomous indicators. For clustering observations, the K-means method was used with the setting of measuring the minimum distances. The results of the study showed statistically significant relationships (p < 0.05) between indicators of educational activity and student feedback, which made it possible to use these indicators to classify reviews. The study using the cluster analysis made it possible to obtain the optimal number of clusters for solving the task of classifying and confirming the adequacy of the developed methodology. Thus, the correlation analysis showed a direct moderate relationship between irrelevant reviews obtained by the methodology and three-cluster classification at r = 0.70, and Cramer's V test showed a strong association with the results of three-cluster classification (0.70). Using the classification results as a training sample, a logistic regression model was built with an adequacy of 94.97%. The results of the study lead to the conclusion that the obtained methodology and the logistic regression model can be used to clean the initial feedback data from irrelevant reviews based on the indicators of students' learning activities.

Generalizing solutions across functionally similar problems correlates with world knowledge and working memory in 2.5- to 4.5-year-olds

Analogical transfer, denoting the ability to use an action that solved a given problem in order to successfully handle a seemingly different but functionally similar problem, requires well-developed self-regulation, as it draws on previous knowledge and demands selecting and shifting between relevant features while ignoring irrelevant ones. Thus, analogical transfer involves executive functions (EFs), yet the contribution of specific EFs is unclear, particularly during the development of the capacity before the age of 5. Here, for the first time, we investigated the contribution of world knowledge, working memory and set-shifting in 2.5- to 4.5-year-olds’ (N = 86) capacity for single-event analogical transfer in a simple, non-verbal, tool-use task. Analogical transfer was independent of age but was predicted by a measure of world knowledge and a measure of working memory across the age-span tested. Our results suggest that world knowledge and working memory underscore analogical transfer early in development.

Foreground-guided textural-focused person re-identification

Person re-identification (Re-ID) is challenging in computer vision. It is crucial to strengthen discriminative features and suppress irrelevant ones for high-performance. Existing Re-ID approaches have won significant progress by attention mechanisms or introducing key-point/part priors. However, these methods suffer the high cost of building models. Besides, potential feature false alarm problems may happen due to background interference and inaccurate priors. Moreover, they usually ignore the degradation of texture clues which are identity sensitive. In this paper, we propose the Foreground-Guided Textural-Focused Network (FTN) to address these problems. Specifically, our FTN is an end-to-end Re-ID framework, which consists of a Semantic Encoder (S-Enc), a Compact Foreground Attention (CFA) module, and a Texture-Focused Decoder (TF-Dec). First, based on CFA and 2D Gaussian kernel, a coarse foreground-guided learning strategy is developed to suppress the feature false alarms at the source. Its core idea lies in constructing foreground guidance, which forces S-Enc to pay more attention to person instance-level features. Second, the TF-Dec is designed as a lightweight reconstruction task. It is trained via a novel gradient loss, and further forces S-Enc to maintain texture-wise details. Our method is computationally efficient as TF-Dec is abandoned in the inference phase. Extensive experiments are conducted on three large-scale Re-ID datasets Market1501, CUHK03, MSMT17, and two other occluded datasets. The results indicate that FTN achieves superior performance against the state-of-the-art methods, e.g., Rank-1 of 96.2% and 59.0% on Market1501 and Occluded-Duke respectively.

Feature selection for measurement models

PurposeIn the determination of the measurement uncertainty, the GUM procedure requires the building of a measurement model that establishes a functional relationship between the measurand and all influencing quantities. Since the effort of modelling as well as quantifying the measurement uncertainties depend on the number of influencing quantities considered, the aim of this study is to determine relevant influencing quantities and to remove irrelevant ones from the dataset.Design/methodology/approachIn this work, it was investigated whether the effort of modelling for the determination of measurement uncertainty can be reduced by the use of feature selection (FS) methods. For this purpose, 9 different FS methods were tested on 16 artificial test datasets, whose properties (number of data points, number of features, complexity, features with low influence and redundant features) were varied via a design of experiments.FindingsBased on a success metric, the stability, universality and complexity of the method, two FS methods could be identified that reliably identify relevant and irrelevant influencing quantities for a measurement model.Originality/valueFor the first time, FS methods were applied to datasets with properties of classical measurement processes. The simulation-based results serve as a basis for further research in the field of FS for measurement models. The identified algorithms will be applied to real measurement processes in the future.

Challenges of using ICT in the library of an Indian Deemed University

The electronic resources are such products which a library provides in a digital platform. WWW (World Wide Web) is very familiar to the society to search and retrieve useful data, information and knowledge in ICT environment. It is used to find authentic information from all primary resources including statistical information and all parts of educational platforms available on national and international level. ICT influences general policies, comments, feedbacks and opinions on available information in LIS field. Information can be collected from one or more sources and also distributed to one or more audiences. This sometimes involves those who have a stake in, or a right to that information. Information Science generates various terms such as: Information seeking, Information retrieval and Information behavior. It is decided to meet all the library professionals working in the University to elicit required information for the purpose. A structured questionnaire with both close and open ended options was administered to collect information from nine (9) independent libraries under this University and the respondents relevant for the study in the light of their observations and experiences. Maximizing the retrieval of relevant records and minimizing the retrieval of irrelevant ones is often described as maximizing both Recall and Precision. The institutional library should organize orientation workshops and seminars for their users to up skill them in utilizing library resources to the optimum and services will be enhanced.

Validation of the European Quality of Life-5 Dimensions, 3-Level Version (EQ-5D-3L) in Patients With Osteoporotic Vertebral Fracture and Comparing Their Health-Related Quality of Life With That of the General Population in Korea.

ObjectiveThis study examined the validity of the European Quality of Life-5 dimensions, 3-Level version (EQ-5D-3L) in patients with osteoporotic vertebral compression fractures (OVCF) and compared the health-related quality of life (HRQOL) of these patients with that of the general population in Korea.MethodsStudy participants completed several questionnaires, such as the EQ-5D-3L, 12-item short-form health survey version 2.0 (SF-12v2), Roland-Morris disability questionnaire (RMQ), and Pain Numeric Rating Scale (Pain NRS). Spearman’s correlation analyses of the questionnaire results were conducted to examine the discriminant and convergent validity of the EQ-5D-3L. Finally, EQ-5D-3L results from study participants were compared with those from the comparison group of the general population that were match-sampled according to age and sex from the 2013 Korean National Health and Nutritional Examination Survey.ResultsSpearman’s correlation coefficients between EQ-5D-3L and pain NRS, SF-12v2, and RMQ were statistically significant. The coefficients between the relevant dimensions, such as mobility in the EQ-5D-3L and physical functioning in the SF-12v2, were higher than those between irrelevant ones. Problems reported by study participants in each dimension of EQ-5D-3L showed statistically significant lower scores in pain NRS, PCS, and MCS from the SF-12v2 and RMQ scores. In comparison with the general population, the problems reported in the patient group were statistically higher than those reported by the general population.ConclusionThis study showed that EQ-5D-3L will be useful for patients with OVCF in Korea because it could be valid. The HRQOL of patients with OVCF was very low in all dimensions of the EQ-5D-3L.

Multivariate Feature Ranking With High-Dimensional Data for Classification Tasks

In many machine learning classification problems, datasets are usually of high dimensionality and therefore require efficient and effective methods for identifying the relative importance of their attributes, eliminating the redundant and irrelevant ones. Due to the huge size of the search space of the possible solutions, the attribute subset evaluation feature selection methods are not very suitable, so in these scenarios feature ranking methods are used. Most of the feature ranking methods described in the literature are univariate methods, which do not detect interactions between factors. In this paper, we propose two new multivariate feature ranking methods based on pairwise correlation and pairwise consistency, which have been applied for cancer gene expression and genotype-tissue expression classification tasks using public datasets. We statistically proved that the proposed methods outperform the state-of-the-art feature ranking methods <i>Clustering Variation, Chi Squared, Correlation, Information Gain, ReliefF</i> and <i>Significance</i>, as well as other feature selection methods for attribute subset evaluation based on correlation and consistency with the multi-objective evolutionary search strategy, and with the embedded feature selection methods <i>C4.5</i> and <i>LASSO</i>. The proposed methods have been implemented on the WEKA platform for public use, making all the results reported in this paper repeatable and replicable.

Improved Reptile Search Optimization Algorithm Using Chaotic Map and Simulated Annealing for Feature Selection in Medical Field

The increased volume of medical datasets has produced high dimensional features, negatively affecting machine learning (ML) classifiers. In ML, the feature selection process is fundamental for selecting the most relevant features and reducing redundant and irrelevant ones. The optimization algorithms demonstrate its capability to solve feature selection problems. Reptile Search Algorithm (RSA) is a new nature-inspired optimization algorithm that stimulates Crocodiles’ encircling and hunting behavior. The unique search of the RSA algorithm obtains promising results compared to other optimization algorithms. However, when applied to high-dimensional feature selection problems, RSA suffers from population diversity and local optima limitations. An improved metaheuristic optimizer, namely the Improved Reptile Search Algorithm (IRSA), is proposed to overcome these limitations and adapt the RSA to solve the feature selection problem. Two main improvements adding value to the standard RSA; the first improvement is to apply the chaos theory at the initialization phase of RSA to enhance its exploration capabilities in the search space. The second improvement is to combine the Simulated Annealing (SA) algorithm with the exploitation search to avoid the local optima problem. The IRSA performance was evaluated over 20 medical benchmark datasets from the UCI machine learning repository. Also, IRSA is compared with the standard RSA and state-of-the-art optimization algorithms, including Particle Swarm Optimization (PSO), Genetic Algorithm (GA), Grasshopper Optimization algorithm (GOA) and Slime Mould Optimization (SMO). The evaluation metrics include the number of selected features, classification accuracy, fitness value, Wilcoxon statistical test (p-value), and convergence curve. Based on the results obtained, IRSA confirmed its superiority over the original RSA algorithm and other optimized algorithms on the majority of the medical datasets.

Feature Selection: A Review and Comparative Study

Feature selection (FS) is an important research topic in the area of data mining and machine learning. FS aims at dealing with the high dimensionality problem. It is the process of selecting the relevant features and removing the irrelevant, redundant and noisy ones, intending to obtain the best performing subset of original features without any transformation. This paper provides a comprehensive review of FS literature intending to supplement insights and recommendations to help readers. Moreover, an empirical study of six well-known feature selection methods is presented so as to critically analyzing their applicability.

Assessment of Sentiment Analysis Using Information Gain Based Feature Selection Approach

Sentiment analysis is the process of determining the intention or emotion behind an article. The subjective information from the context is analyzed by the sentimental analysis of the people’s opinion. The data that is analyzed quantifies the reactions or sentiments and reveals the information’s contextual polarity. In social behavior, sentiment can be thought of as a latent variable. Measuring and comprehending this behavior could help us to better understand the social issues. Because sentiments are domain specific, sentimental analysis in a specific context is critical in any real-world scenario. Textual sentiment analysis is done in sentence, document level and feature levels. This work introduces a new Information Gain based Feature Selection (IGbFS) algorithm for selecting highly correlated features eliminating irrelevant and redundant ones. Extensive textual sentiment analysis on sentence, document and feature levels are performed by exploiting the proposed Information Gain based Feature Selection algorithm. The analysis is done based on the datasets from Cornell and Kaggle repositories. When compared to existing baseline classifiers, the suggested Information Gain based classifier resulted in an increased accuracy of 96% for document, 97.4% for sentence and 98.5% for feature levels respectively. Also, the proposed method is tested with IMDB, Yelp 2013 and Yelp 2014 datasets. Experimental results for these high dimensional datasets give increased accuracy of 95%, 96% and 98% for the proposed Information Gain based classifier for document, sentence and feature levels respectively compared to existing baseline classifiers.

Effect of sustained selective attention on steady-state visual evoked potentials.

Visual-spatial selective attention enhances the processing of task-relevant visual events while suppressing the processing of irrelevant ones. In this study, we employed a frequency-tagging paradigm to investigate how sustained visual-spatial attention modulates the first harmonic and second harmonic steady-state visual evoked potentials (SSVEPs). Unlike previous studies, that investigated stimulation durations of 10s or less, we tested a 30-s period. SSVEPs were elicited by simultaneously presenting to the right and left visual hemifields two pattern reversal checkerboard stimuli modulating at 7.14Hz and 11.11Hz. Participants were cued to selectively attend to one visual hemifield while ignoring the other. Behavioral results indicated that participants selectively attended to the cued visual hemifield. When participants attended to the visual stimuli, there were larger second harmonic SSVEPs but no attentional modulation of first harmonics. The results are consistent with the proposal that neural populations underlying first, and second harmonics have distinct functional roles, i.e., first harmonics' mechanisms preserve stimulus properties and are resistant to attentional gain, whereas second harmonics mediate attentional modulation. This interpretation is supported by a gain control theory of selective attention.

Modulating Cortical Instrument Representations During Auditory Stream Segregation and Integration With Polyphonic Music.

Numerous neuroimaging studies demonstrated that the auditory cortex tracks ongoing speech and that, in multi-speaker environments, tracking of the attended speaker is enhanced compared to the other irrelevant speakers. In contrast to speech, multi-instrument music can be appreciated by attending not only on its individual entities (i.e., segregation) but also on multiple instruments simultaneously (i.e., integration). We investigated the neural correlates of these two modes of music listening using electroencephalography (EEG) and sound envelope tracking. To this end, we presented uniquely composed music pieces played by two instruments, a bassoon and a cello, in combination with a previously validated music auditory scene analysis behavioral paradigm (Disbergen et al., 2018). Similar to results obtained through selective listening tasks for speech, relevant instruments could be reconstructed better than irrelevant ones during the segregation task. A delay-specific analysis showed higher reconstruction for the relevant instrument during a middle-latency window for both the bassoon and cello and during a late window for the bassoon. During the integration task, we did not observe significant attentional modulation when reconstructing the overall music envelope. Subsequent analyses indicated that this null result might be due to the heterogeneous strategies listeners employ during the integration task. Overall, our results suggest that subsequent to a common processing stage, top-down modulations consistently enhance the relevant instrument’s representation during an instrument segregation task, whereas such an enhancement is not observed during an instrument integration task. These findings extend previous results from speech tracking to the tracking of multi-instrument music and, furthermore, inform current theories on polyphonic music perception.

Unitization congruence moderates the effect of unitization on associative recognition: evidence from event-related potentials

Recently, a growing number of studies have shown that familiarity can promote associative recognition when more than two types of stimuli are combined into a single entity. However, it is not clear why the effect of unitization on associative recognition varies in the different studies. In this study, we examined whether the unitization congruence (UC) between original image pairs and rearranged ones affects associative recognition. Participants were asked to learn relevant and irrelevant picture pairs for the encoding phase, and they needed to distinguish intact pairs from rearranged pairs for the retrieval phase, while event-related potentials (ERPs) were recorded during the whole process. The results of the behavioral experiments showed that the performances of recognition (Prs: hits-false alarms) were higher under the relevant condition than under the irrelevant one for both UC conditions, although this effect was much larger under the inconsistent UC condition than under the consistent one. The ERP results also revealed that there were larger early old/new effect (FN400) and late left parietal old/new effect (LPC) under the relevant condition than under the irrelevant ones, irrespective of its consistency or inconsistency. However, these effects were more significant for inconsistent conditions. These results indicated that UC can moderate the effect of unitization on associative recognition by increasing familiarity and recollection and revealed the need to match the level of unitization of materials arrangements in future studies.

The Role of Exercise and Physical Activity against COVID-19: A Narrative Review Study

The aim of this study was to understand the role of sport on covid-19 through reviewing the published literature. This was a narrative review that used valid keywords for searching the related scholarly articles. The search for the related keywords on Pubmed, Scopus, Science Direct library and Google Scholar lead to 1158 articles in 2020. After screening 28 article found eligibility to be inclouded in the review. The result indicated that physical activity and exercise play a role in preventing novel coronavirous infection and strengthening the immune system againt the covid-19. Extended Abstract Background and Objectives Since COVID-19 virus is an unknown virus which has caused alarming death toll over the past year, researchers felt implelled to take measures through doing extensive research studies in a variety of related fields to control the virus. On the other hand, exercise is known as a useful non-pharmacological method to improve the function of the immune system, so researchers in the field of sports science have done valuable research studies in this research domain, albeit limited, which can provide us with valuable information in this regard. Hence, there is a pressing need to collect and categorize these latest findings. Therefore, the very aim of this study was to review literature on the role of exercise and physical activity against the COVID-19. Material and Methods This was a narrative review, the data was collected through searching and reteriving available articles published in information resources like Medline (PubMed), Scopus , Science Direct collection, and also Google Scholar. The keywords employed for search strategy building included: "Exercise Training", "Physical Activity", "Aerobic Exercise", "Interval Training", "Resistance Training", "Endurance Training", "COVID-19", "novel Coronavirus" and "Immune System". Based on the results of our searches conducted in databases, initially 1185 articles were selected. After screening the titles and abstracts of the retrieved articles and removal of the duplicate and irrelevant ones, 59 articles were included in the reveiew stage. Then, full text of eligible articles were put underclose scrutiny which led to the final selection of only 28 articles to be studied. Results According to our review, exercise can play a role in different ways against COVID-19 diseases and novel coronavirous, which is now recognized as a global problem. These roles include preventing some underlying diseases such as hypertension, coronary artery disease and diabetes and improving immunity. Exercise lowers cortisol levels in long term, which reduces the stress induced by COVID-19, which in turn strengthens the immune system. On the other hand, doing moderate- intensity exercise can strengthen the immune system in the long run, while low-intensity or high-intensity exercise types do not bring about similar degree of adaption. Conclusion It is necessary to pay special attention to exercise training in Covid-19 pandemic period. However, it should be noted that due to the novelty of this phenomenon, limited research has been done in this area to date and more research studies are required to be done in this regard. Practical implications of research It is best for people to stay home during Covid-19 pandemic period and to have 150 minutes of moderate-intensity regular exercise during the week to boost their immune system. Ethical considerations In this review research, the ethical issues of research and fiduciary have been observed. Conflict of interest We have no conflicts of interest to disclose. Acknowledgment No sources of funding were used to assist in the preparation of this article.

Economic Dependencies in Integrated Circuits

Life can never deliver too many reminders to approach forecasting with humility. The recent shortages in custom chips delivered another such reminder. There also may be a more epistemological lesson to learn from these events. Good analysis directs attention at the right facts, beliefs, or opinions, and dismisses irrelevant ones. Perhaps the wrong aspects received too much attention and the most relevant ones were dismissed. What did most analysts misunderstand that led to not forecasting these shortages? In other words, let us reconsider whether the prevailing view of this market analyzes events in the terms most relevant to understanding them. Today’s column argues that these shortages illustrate the role of economic dependence in the face of demand volatility. Perhaps those topics ought to be elevated to a central question in industry analysis.

Consumer‐led approach to adapt a food‐odors emotional lexicon for the Spanish population: A tool for designing the scent of food spaces

AbstractAromas are powerful stimuli capable of triggering emotions. The Geneva Emotion and Odor Scale (GEOS) was developed to measure the affective response to odors, and was later adapted to other countries and cultures (EOSs). Also, an universal scale was proposed (UniGEOS) to be used regardless of the culture, but without considering Spanish‐speaking cultures. During the present research, a lexicon for measuring the emotions evoked by food‐related odors has been developed, adapting the aforementioned scales to the Spanish population and using a consumer‐led approach. Using EOSs lists as starting point, the procedure was: to discard nonemotional terms and translate the selected ones, to group the emotions and remove the irrelevant ones for the food‐odors category, and to test the lexicon. The developed lexicon (SEFrOS) contained six categories, 25 terms. The lexicon shared some dimensions with EOSs, but with specific differences, maybe related to culture or the food‐linked nature of the lexicon.Practical applicationsEmotions elicited by foods have been investigated during the last years to favor proper products positioning in the market. Odors have been used to modulate customer perceptions and attitudes towards products and stores, because of being closely linked to these emotions' elicitation. Although some lexicons have been developed to assess food‐related emotions, no lexicon has been developed to specifically evaluate the emotions elicited by food‐odors, and to study the emotions elicited by the odor of food‐related spaces. This research presents the development of a new lexicon for food‐odors, using previously validated odor scales as starting point, and adapting them to the Spanish population, well‐known because of its eating‐out model closely related to different food spaces.

EBCR: Empirical Bayes concordance ratio method to improve similarity measurement in memory-based collaborative filtering.

Recommender systems aim to provide users with a selection of items, based on predicting their preferences for items they have not yet rated, thus helping them filter out irrelevant ones from a large product catalogue. Collaborative filtering is a widely used mechanism to predict a particular user’s interest in a given item, based on feedback from neighbour users with similar tastes. The way the user’s neighbourhood is identified has a significant impact on prediction accuracy. Most methods estimate user proximity from ratings they assigned to co-rated items, regardless of their number. This paper introduces a similarity adjustment taking into account the number of co-ratings. The proposed method is based on a concordance ratio representing the probability that two users share the same taste for a new item. The probabilities are further adjusted by using the Empirical Bayes inference method before being used to weight similarities. The proposed approach improves existing similarity measures without increasing time complexity and the adjustment can be combined with all existing similarity measures. Experiments conducted on benchmark datasets confirmed that the proposed method systematically improved the recommender system’s prediction accuracy performance for all considered similarity measures.

The role of component-resolved diagnosis in Hymenoptera venom allergy in clinical practice

Background: Hymenoptera venom allergy is an immunoglobulin (Ig) E mediated hypersensitivity reaction to Hymenoptera venoms. Obvious identification of the culprit insect that causes the clinical symptoms and, hence, the accurate selection of venom for curative treatment, is of great importance for the effectiveness and safety of venom immunotherapy. Objective: In this study, the contribution of component-resolved diagnostics (CRD) is evaluated in the diagnosis of Hymenoptera venom allergy. Method: Ninety-three patients from four different centers in Turkey were included in the study. Conventional tests, including prick and intradermal skin tests, with commercial venom extracts and serum specific IgE (sIgE) levels for whole venoms were performed. An sIgE analysis for venom allergen components, including rApi m 1, rApi m 2, rApi m 10, rVes v 1, rVes v 5, were evaluated by immunoblotting. Results: In conventional test results, 17 of 35 patients with bee venom allergy were positive to honey bee venom, whereas 18 patients were positive to bee and wasp venoms. In 28 of 35 patients with bee venom allergy, the diagnosis was confirmed with CRD. CRD revealed a sensitivity of 80% in patients with bee venom allergy. According to conventional tests, 7 of 24 patients with vespid venom allergy demonstrated sensitivity only to Vespula species, whereas 17 patients revealed double positivity. The total diagnostic sensitivity of Ves v 1 and Ves v 5 was calculated as 87.5%. Ten of 23 patients with a history of hypersensitivity to both venoms showed double sensitivity with CRD; one patient had cross-reactivity, one patient was found to be sensitive only to bee venom, and, eight patients were sensitive only to Vespula species. Eleven patients had an uncertain history in terms of the culprit insect type and six of them had double sensitivity in CRD. Conclusion: CRD seemed to be more helpful in diagnosing vespid venom allergy than bee venom allergy. It can also discriminate clinically significant sensitizations from irrelevant ones.

A new feature selection metric for text classification: eliminating the need for a separate pruning stage

Terms that occur too frequently or rarely in various texts are not useful for text classification. Pruning can be used to remove such irrelevant terms reducing the dimensionality of the feature space and, thus making feature selection more efficient and effective. Normally, pruning is achieved by manually setting threshold values. However, incorrect threshold values can result in the loss of many useful terms or retention of irrelevant ones. Existing feature ranking metrics can assign higher ranks to these irrelevant terms, thus degrading the performance of a text classifier. In this paper, we propose a new feature ranking metric, which can select the most useful terms in the presence of these too frequently and rarely occurring terms, thus eliminating the need for pruning these terms. To investigate the usefulness of the proposed metric, we compare it against seven well-known feature selection metrics on five data sets namely Reuters-21578 (re0, re1, r8) and WebACE (k1a, k1b) using multinomial naive Bayes and support vector machines classifiers. Our results based on a paired t-test show that the performance of our metric is statistically significant than that of the other seven metrics.