A bi-objective non-linear approach for determining the ordering strategy for group B in ABC analysis inventory

The aim of the present study was to examine Turkish teacher candidates’ competency levels in writing different types of test items by utilizing Rasch analysis. In addition, the effect of the expertise of the raters scoring the items written by the teacher candidates was examined within the scope of the study. 84 Turkish teacher candidates participated in the present study, which was conducted using the relational survey model, one of the quantitative research methods. Three experts participated in the rating process: an expert in Turkish education, an expert in measurement and evaluation, and an expert in both Turkish education and measurement and evaluation. The teacher candidates wrote true-false, short response, multiple choice and open-ended types of items in accordance with the Test Item Development Form, and the raters scored each item type by designating a score between 1 and 5 based on the item evaluation scoring rubric prepared for each item type. The study revealed that Turkish teacher candidates had the highest level of competency in writing true-false items, while they had the lowest competency in writing multiple-choice items. Moreover, it was revealed that raters’ expertise had an effect on teacher candidates’ competencies in writing different types of items. Finally, it was found that the rater who was an expert in both Turkish education and measurement and evaluation had the highest level of scoring reliability, while the rater who solely had expertise in measurement and evaluation had the relatively lowest level of scoring reliability.

This paper looks at a question that has up to now not yet been put, namely which textual dictionary structure elements with which features, by the use of which the user can find semantic and pragmatic knowledge within an article can systematically be differentiated. In order to do this the following four dictionary structure elements are introduced: items, item texts, functional item additions and item symbols. Then the dictionary structure elements for enabling knowledge regarding meaning are treated by means of examples, e.g. different types of items giving the meaning, different types of items giving semantically related expressions, different types of pragmatically enriched items giving the paraphrase of meaning, etc. Finally, the dictionary structure elements enabling pragmatic knowledge are discussed, e.g. different types of diasystematic labelling items and pragmatic labelling symbols. In conclusion, a brief perspective is given on the relevant approach to meaning.

Given the success of online shopping platforms and e-commerce technology, there is an increasing necessity to quickly and safely package different types of items. Addressing this necessity requires technology to accurately measure items at a high speed. Existing studies, however, lack in terms of reproducibility and the diversity of the items measured. In this paper, we present a novel approach for item measurement, targeting automated packaging systems that make use of belt conveyors. In particular, leveraging a scenario-driven approach and an automata-based control design, we describe in detail the creation of a real-world prototype for belt conveyor-based item measurement. Experimental results obtained for this prototype demonstrate that it is possible to measure different types of items (boxes, books, household items) with a mean error of less than 3.02mm and a standard deviation of less than 2.34mm, for a maximum conveyor belt speed that is less than 0. 5m /s and a maximum calculation time of 20ms.

Human beings are creatures of habit. In their daily life, people tend to repeatedly consume similar types of food items over several days and occasionally switch to consuming different types of items when the consumptions become overly monotonous. However, the novel and repeat consumption behaviors have not been studied in food recommendation research. More importantly, the ability to predict daily eating habits of individuals is crucial to improve the effectiveness of food recommender systems in facilitating healthy lifestyle change. In this study, we analyze the patterns of repeat food consumptions using large-scale consumption data from a popular online fitness community called MyFitnessPal (MFP), conduct an offline evaluation of various state-of-the-art algorithms in predicting the next-day food consumption, and analyze their performance across different demographic groups and contexts. The experiment results show that algorithms incorporating the exploration-and-exploitation and temporal dynamics are more effective in the next-day recommendation task than most state-of-the-art algorithms.

Dynamic scheduling for production systems operating in a random environment

Examining the Language Factor in Mathematics Assessments Adnan Kan, Okan Bulut Abstract In educational testing, assessment specialists typically create multiple forms of a test for different purposes, such as increasing test security or developing an item bank. Using different types of items across test forms is also a common practice to create alternative test forms.This study investigates whether word problems and mathematically expresseditems can be used interchangeably regardless of their linguistic complexities. A sample of sixth grade students was given two forms of a mathematics assessment. The first form included mathematicsitems based onmathematical terms, expressions, and equations. The second form included the same items as word problems. The underlying tasks and solutions of the items in the first test form were the same as the corresponding items in the second form. Explanatory item response modeling was used for examining the impact of item type and genderon difficulty levels of items and students’ test scores. The results showed that word problems were easier than mathematically expressed items. Gender and its interaction with the linguistic complexity of mathematics items did not seem to have any impact on student performance on the test. Full Text: PDF DOI: 10.15640/jehd.v4n1a13

It is assumed that a population consists of r(r < ∞) types of items, each item being characterized by a real-valued, non-negative random variable called the item's amount. Each type of item may have a different distribution of amounts. It is also assumed that samples are selected from this population so as to contain a fixed total amount rather than a fixed number of items. Finding the asymptotic distribution of the sample amounts for the different types of items, estimating the population amount proportions, and estimating the average item amount are three problems for which answers are found by use of methods and theorems from renewal theory. Some results which have already appeared in the literature are derived and extended in the course of this work.

Manufacturing companies are using collaborative planning for the coordination of lot-sizing decisions in inter-organisational supply chains. By using collaborative planning, the members of a supply chain try to identify a production plan which results in lower costs compared to individual plans by simultaneously preserving their autonomy. A distributed lot-sizing problem with rivaling agents (DULR) is studied where each item can be produced by more than one member of the coalition (agent). Thereby it occurs that agents compete for the production quotas of items. However, the goal of this contribution is to extend the DULR by considering two types of items. One type can be produced by more than one agent, while the other one can only be produced by a certain agent due to contractual obligations. We denote the former type of items as concurrent item and the latter one as compulsory items. To solve the DULR with different types of items, an existing negotiation mechanism based on a simulated annealing is applied and modified. A benchmark study shows that the modified solution approach even outperforms the best-known approach for the DULR. Based on this finding, a second study is applied where the impact of compulsory items is investigated for the DULR.

Recommendation system has been a widely studied task both in academia and industry. Previous works mainly focus on homogeneous recommendation and little progress has been made for heterogeneous recommender systems. However, heterogeneous recommendations, e.g., recommending different types of items including products, videos, celebrity shopping notes, among many others, are dominant nowadays. State-of-the-art methods are incapable of leveraging attributes from different types of items and thus suffer from data sparsity problems. And it is indeed quite challenging to represent items with different feature spaces jointly. To tackle this problem, we propose a kernel-based neural network, namely deep unified representation (or DURation) for heterogeneous recommendation, to jointly model unified representations of heterogeneous items while preserving their original feature space topology structures. Theoretically, we prove the representation ability of the proposed model. Besides, we conduct extensive experiments on real-world datasets. Experimental results demonstrate that with the unified representation, our model achieves remarkable improvement (e.g., 4.1% ~ 34.9% lift by AUC score and 3.7% lift by online CTR) over existing state-of-the-art models.

A psychological framework for different types of items commonly used with mixed-format exams is proposed. A choice model based on signal detection theory (SDT) is used for multiple-choice (MC) items, whereas an item response theory (IRT) model is used for open-ended (OE) items. The SDT and IRT models are shown to share a common conceptualization in terms of latent states of "know/don't know" at the examinee level. This in turn suggests a way to join or "fuse" the models-through the probability of knowing. A general model that fuses the SDT choice model, for MC items, with a generalized sequential logit model, for OE items, is introduced. Fitting SDT and IRT models simultaneously allows one to examine possible differences in psychological processes across the different types of items, to examine the effects of covariates in both models simultaneously, to allow for relations among the model parameters, and likely offers potential estimation benefits. The utility of the approach is illustrated with MC and OE items from large-scale international exams.

Traditionally, the effect of assessment item types including true/false questions (TFQs), multiple-choice questions (MCQs), short answer questions (SAQs), and case scenario questions (CSQs) is examined through psychometric qualities or student interviews. However, brain activity while answering such questions or items remains unknown. Functional near-infrared spectroscopy (fNIRS) can be used to safely measure cerebral cortex hemodynamic response during various tasks. Hence, this fNIRS study aimed to determine differences in frontotemporal cortex activity as medical students answered TFQs, MCQs, SAQs, and CSQs. In total, 24 medical students (13 males and 11 females) were recruited in this study during their mid-psychiatry posting. Oxy-hemoglobin and deoxy-hemoglobin levels in the frontal and temporal regions were measured with a 52-channel fNIRS system. Participants answered 9-18 trials under each of the four types of tasks that were based on their psychiatry curriculum during fNIRS measurements. The area under the oxy-hemoglobin curve (AUC) for each participant and each item type was derived. Repeated measures ANOVA with post-hoc Bonferroni-corrected pairwise comparisons were used to determine differences in oxy-hemoglobin AUC between TFQs, MCQs, SAQs, and CSQs. Oxy-hemoglobin AUC was highest during the CSQs, followed by SAQs, MCQs, and TFQs in both the frontal and temporal regions. Statistically significant differences between different types of items were observed in oxy-hemoglobin AUC of the frontal region (p ≤ 0.001). Oxy-hemoglobin AUC in the frontal region was significantly higher during the CSQs than TFQ (p = 0.005) and during the SAQ than TFQ (p = 0.025). Although the percentage of correct responses was significantly lower in MCQ than in the other item types, there was no correlation between the percentage of correct response and oxy-hemoglobin AUC in both regions for all four item types (p > 0.05). CSQs and SAQs elicited greater hemodynamic response than MCQs and TFQs in the prefrontal cortex of medical students. This suggests that more cognitive skills may be required to answer CSQs and SAQs.

Quasi-Newton’s method for multiobjective optimization

We consider a matching system with random arrivals of items of different types. The items wait in queues—one per item type—until they are “matched.” Each matching requires certain quantities of items of different types; after a matching is activated, the associated items leave the system. There exists a finite set of possible matchings, each producing a certain amount of “reward.” This model has a broad range of important applications, including assemble-to-order systems, Internet advertising, and matching web portals. We propose an optimal matching scheme in the sense that it asymptotically maximizes the long-term average matching reward, while keeping the queues stable. The scheme makes matching decisions in a specially constructed virtual system, which in turn controls decisions in the physical system. The key feature of the virtual system is that, unlike the physical one, it allows the queues to become negative. The matchings in the virtual system are controlled by an extended version of the greedy primal–dual (GPD) algorithm, which we prove to be asymptotically optimal—this in turn implies the asymptotic optimality of the entire scheme. The scheme is real time; at any time, it uses simple rules based on the current state of the virtual and physical queues. It is very robust in that it does not require any knowledge of the item arrival rates and automatically adapts to changing rates. The extended GPD algorithm and its asymptotic optimality apply to a quite general queueing network framework, not limited to matching problems, and therefore are of independent interest.

Recommender Systems ( RecSys ) provide suggestions in many decision-making processes. Given that groups of people can perform many real-world activities (e.g., a group of people attending a conference looking for a place to dine), the need for recommendations for groups has increased. A wide range of Group Recommender Systems ( GRecSys ) has been developed to aggregate individual preferences to group preferences. We analyze 175 studies related to GRecSys . Previous works evaluate their systems using different types of groups (sizes and cohesiveness), and most of such works focus on testing their systems using only one type of item, called Experience Goods (EG). As a consequence, it is hard to get consistent conclusions about the performance of GRecSys . We present the aggregation strategies and aggregation functions that GRecSys commonly use to aggregate group members’ preferences. This study experimentally compares the performance (i.e., accuracy, ranking quality, and usefulness) using four metrics (Hit Ratio, Normalize Discounted Cumulative Gain, Diversity, and Coverage) of eight representative RecSys for group recommendations on ephemeral groups. Moreover, we use two different aggregation strategies, 10 different aggregation functions, and two different types of items on two types of datasets (EG and Search Goods (SG)) containing real-life datasets. The results show that the evaluation of GRecSys needs to use both EG and SG types of data, because the different characteristics of datasets lead to different performance. GRecSys using Singular Value Decomposition or Neural Collaborative Filtering methods work better than others. It is observed that the Average aggregation function is the one that produces better results.

It is assumed that a population consists of two different types of items, each item being characterized by a random variable called the item's amount. Each type of item may have a different distribution of amounts. Samples are selected from this population so as to contain a fixed total amount rather than a fixed number of items. Some asymptotic properties of the sample amount proportions are derived for the case in which the sampled items have Markov dependence property with respect to type. Estimation of the asymptopic variance is discussed and an application to line of areas is given.