Unobserved Categories Research Articles

CHANGES IN THE PERCEPTION OF TELEMATICS TECHNOLOGY BY ROAD TRANSPORT COMPANIES: AN EMPIRICAL ANALYSIS IN 2020-21

We present a novel study concerning the attitudes of road transport enterprises towards a broad application of telematics in operational management in road transportation. The study aims to assess telematics application in road transport and its changes over time while showing the factors most likely to determine the systems’ use. Unobserved categories defined in the technology acceptance model (TAM) are adjusted to measure perceived usefulness, perceived ease of use, and attitudes toward using telematics systems by road transport managers. The study is based on 323 transport enterprises analyzed in two waves in 2020 and 2021. The use of two different time points is motivated by an observed increase in the digitalization of transport documents caused by the COVID-19 pandemic. The empirical findings support the TAM’s usefulness in evaluating IT in transport business management. The findings also reveal that the significantly increased telematics use in 2020 was observed while it was endured. The results are checked for robustness and used for simulations. The study compares managers’ behaviors over time and simulates the effect of individual (observed) variables on unobserved TAM categories.

Semi-Supervised Low-Rank Semantics Grouping for Zero-Shot Learning.

Zero-shot learning has received great interest in visual recognition community. It aims to classify new unobserved classes based on the model learned from observed classes. Most zero-shot learning methods require pre-provided semantic attributes as the mid-level information to discover the intrinsic relationship between observed and unobserved categories. However, it is impractical to annotate the enriched label information of the observed objects in real-world applications, which would extremely hurt the performance of zero-shot learning with limited labeled seen data. To overcome this obstacle, we develop a Low-rank Semantics Grouping (LSG) method for zero-shot learning in a semi-supervised fashion, which attempts to jointly uncover the intrinsic relationship across visual and semantic information and recover the missing label information from seen classes. Specifically, the visual-semantic encoder is utilized as projection model, low-rank semantic grouping scheme is explored to capture the intrinsic attributes correlations and a Laplacian graph is constructed from the visual features to guide the label propagation from labeled instances to unlabeled ones. Experiments have been conducted on several standard zero-shot learning benchmarks, which demonstrate the efficiency of the proposed method by comparing with state-of-the-art methods. Our model is robust to different levels of missing label settings. Also visualized results prove that the LSG can distinguish the test unseen classes more discriminative.

Self-perceptions of English and Welsh constables and sergeants preparedness for online crime

PurposeCybercrime is the greatest threat facing law enforcement agencies in England and Wales. Although these crimes are transnational by nature, the burden of response has been placed on line officers. Not all officers, however, believe they are capable of responding to calls involving cybercrime. The current study, using latent class analysis (LCA) on a large sample of English and Welsh officers, finds two types of officers: those prepared (39 percent) and those unprepared (61 percent). Using logistic regression to predict who falls into either classification, the authors find that training and age are the best predictors of latent membership. Implications for policy and future research are discussed. The paper aims to discuss these issues.Design/methodology/approachIn this study, the authors used LCA to determine the number and character of unobserved categories of officers in how they deal with cybercrime.FindingsThe LCA indicated there are two distinct categories of police in the English and Welsh constabulary: those prepared (39 percent) and those unprepared (61 percent). Training and age were the two key determinants of this classification.Research limitations/implicationsLCA is an exploratory analysis technique that requires additional validation to confirm the findings of any one study.Practical implicationsThe salience of training in helping officers feel prepared to deal with cybercrime cases as well as victims was demonstrated. A full 60 percent of the officers in this study fell in the “unprepared” category, which continues to highlight the limitations of local police to handle cybercrime cases; nevertheless, almost 40 percent of officer could be considered ready when responding to cybercrimes.Social implicationsAs the harm cybercrime brings to our financial and social well-being, law enforcement agencies will be required to improve their response capabilities. Most current cybercrime responses address technical issues related to online fraud and abuse, but officers often perceive the problem as outside their legal and geographic jurisdiction. Knowing how officers perceive cybercrime as well as their own capabilities will allow us to begin changing enforcement policies, training capacity and individual response efficacy.Originality/valueThis study involved a sample of English and Welsh constables and sergeants to classify their cybercrime readiness. The analysis and particular data are unique to the study of cybercrime.

A new coefficient of interrater agreement: The challenge of highly unequal category proportions.

We derive a general structure that encompasses important coefficients of interrater agreement such as the S-coefficient, Cohen's kappa, Scott's pi, Fleiss' kappa, Krippendorff's alpha, and Gwet's AC1. We show that these coefficients share the same set of assumptions about rater behavior; they only differ in how the unobserved category proportions are estimated. We incorporate Bayesian estimates of the category proportions and propose a new agreement coefficient with uniform prior beliefs. To correct for guessing in the process of item classification, the new coefficient emphasizes equal category probabilities if the observed frequencies are unstable due to a small sample, and the frequencies increasingly shape the coefficient as they become more stable. The proposed coefficient coincides with the S-coefficient for the hypothetical case of zero items; it converges to Scott's pi, Fleiss' kappa, and Krippendorff's alpha as the number of items increases. We use simulation to show that the proposed coefficient is as good as extant coefficients if the category proportions are equal and that it performs better if the category proportions are substantially unequal. (PsycINFO Database Record (c) 2019 APA, all rights reserved).

A Proposal for Classification of Document Data with Unobserved Categories Considering Latent Topics

A Proposal for Classification of Document Data with Unobserved Categories Considering Latent Topics

A new zero-inflated negative binomial methodology for latent category identification.

We introduce a new statistical procedure for the identification of unobserved categories that vary between individuals and in which objects may span multiple categories. This procedure can be used to analyze data from a proposed sorting task in which individuals may simultaneously assign objects to multiple piles. The results of a synthetic example and a consumer psychology study involving categories of restaurant brands illustrate how the application of the proposed methodology to the new sorting task can account for a variety of categorization phenomena including multiple category memberships and for heterogeneity through individual differences in the saliency of latent category structures.

The Heterogeneous P-Median Problem for Categorization Based Clustering

The p-median offers an alternative to centroid-based clustering algorithms for identifying unobserved categories. However, existing p-median formulations typically require data aggregation into a single proximity matrix, resulting in masked respondent heterogeneity. A proposed three-way formulation of the p-median problem explicitly considers heterogeneity by identifying groups of individual respondents that perceive similar category structures. Three proposed heuristics for the heterogeneous p-median (HPM) are developed and then illustrated in a consumer psychology context using a sample of undergraduate students who performed a sorting task of major U.S. retailers, as well as a through Monte Carlo analysis.

Identifying consumer heterogeneity in unobserved categories

Categorization has been extensively studied in both the psychology and marketing literatures. However, very little methodological research has demonstrated the heterogeneity in consumers’ unobserved category structures and activations. We propose a new latent structure procedure that simultaneously identifies the unobserved categories that consumers use and represents consumer heterogeneity via different groups of consumers who have activated different unobserved latent categories. The results of an empirical study in Sports Marketing about sports fans’ perceptions of various sports illustrates how the proposed methodology can capture heterogeneity at the group level and account for a variety of different category structures.