Search results

The purpose of this paper is to show how data mining techniques can improve the performance management of the judiciary, helping judges in steering position with specific and timely measures. It explores different approaches to analyse the length of trials, based on the case of an Italian judicial office.

The paper presents a temporal analysis to compare the timeliness of trials, using data and process mining approaches with the support of a specific software to represent graphically the results. Data were gathered directly from the office data base, improving precision and the opportunity to monitor specific phases of the trials.

The results highlight the progress that can be reached using data mining approaches to develop performance analyses helping courts to correct inefficiencies and to manage the personnel distribution, overcoming the critical comments arisen against traditional KPI (Raine, 2000). The work proposes a methodology to analyse cases deriving from different juridical matters useful to set up a performance monitoring system that could be diffused to different courts.

The limitations of the research regard the analysis of a selected, limited number of cases in terms of judicial matters.

Data mining techniques can improve the performance management processes in providing more accurate feedback to the judicial offices leaders and increasing the organisational learning.

The performance of the judiciary is one of the relevant issues that emerged in the recent decade in the field of public sector reforms. Several reasons explain this interest, which has gone beyond the specific legal disciplines to involve public policy, management, economics and ICT studies.

Considering the literature on the judiciary (Visser et al., 2019; Di Martino et al., 2021; Troisi and Alfano, 2023) the contribution differs as both the methodological approach and the predictive analysis considers the intrinsic differences that define cases belonging to different juridical matters performing a cross-sectional analysis, with a specific focus of process variants.

Experiments are the backbone of the development process of data-driven predictive models for scientific applications. The quality of the experiments directly impacts the model performance. Uncertainty inherently affects experiment measurements and is often missing in the available data sets due to its estimation cost. For similar reasons, experiments are very few compared to other data sources. Discarding experiments based on the missing uncertainty values would preclude the development of predictive models. Data profiling techniques are fundamental to assess data quality, but some data quality dimensions are challenging to evaluate without knowing the uncertainty. In this context, this paper aims to predict the missing uncertainty of the experiments.

This work presents a methodology to forecast the experiments’ missing uncertainty, given a data set and its ontological description. The approach is based on knowledge graph embeddings and leverages the task of link prediction over a knowledge graph representation of the experiments database. The validity of the methodology is first tested in multiple conditions using synthetic data and then applied to a large data set of experiments in the chemical kinetic domain as a case study.

The analysis results of different test case scenarios suggest that knowledge graph embedding can be used to predict the missing uncertainty of the experiments when there is a hidden relationship between the experiment metadata and the uncertainty values. The link prediction task is also resilient to random noise in the relationship. The knowledge graph embedding outperforms the baseline results if the uncertainty depends upon multiple metadata.

The employment of knowledge graph embedding to predict the missing experimental uncertainty is a novel alternative to the current and more costly techniques in the literature. Such contribution permits a better data quality profiling of scientific repositories and improves the development process of data-driven models based on scientific experiments.

This paper aims to investigate actual tourist customer visiting behavior with behavioral data from Google Popular Times to evaluate the extent that such an online source is useful to better understand, analyze and predict tourist consumer behaviors.

Following six hypotheses on tourist behavior, a purpose-built software tool was developed, pre-tested, and then used to obtain a large-scale data sample of 20,000 time periods for 198 restaurants. Both bi-variate linear regression and correlation analyzes were used for hypothesis testing.

Support was established for the hypotheses, through an analysis of customer reviews, timing effects, the number of pictures uploaded and price segment information provided by tourists to a given restaurant. Also, a relationship to average duration time was found to be positive. The findings demonstrate that data provided through Google Popular Times matches theoretical and logical assumptions to a high degree. Thus, the data source is potentially powerful for providing valuable information to stakeholders (e.g. researchers, managers and tourists).

This paper is the first to both conceptually and empirically demonstrate the practicality and value of Google Popular Times to better understand, analyze and predict tourist consumer behaviors. Value is thereby provided by the potential for this approach to offer insights based behavioral data. Importantly, until now such an approach to gathering and analyzing this volume of actual customer data was previously considered far less practical in terms of time and expense.

本研究旨在用谷歌热门时段(Google Popular Times)的行为数据来探讨游客的实际访问行为, 以评估此种线上资源对理解、分析和预测游客消费行为的实用程度。

基于对游客访问行为的六个假设, 本研究开发并前测一种专用软件工具, 用其收集198家餐厅中20,000个时间段的大规模数据样本。双变量线性回归(bi-variate linear regression)和相关性分析(correlation analyzes)均用于检验假设。

研究结果支持本文假设, 包含顾客评论数量、时间影响、图片数量以及价格区间等资讯對特定餐厅游客平均数量之預測; 同时亦发现与平均持续时间的正向关系。研究结果证明, 谷歌热门时段所提供的数据很大程度上符合理论与逻辑假设。因此, 其具备潜在强大功能, 能为利害关系人（如研究者, 管理者, 游客）提供高价值的资讯。

本研究是第一个从概念与实证上证明谷歌热门时段的实用性和价值, 进而深入理解、分析和预测游客消费行为。此方式透过行为数据来提供深入的见解并创造价值; 重要的是, 在此之前, 这种收集与分析大量实际顾客数据的方法被认为缺乏时间与成本效益。

El presente documento tiene como objetivo investigar las conductas de visita de los clientes turísticos reales con datos de comportamiento de Tiempos populares de Google (Google Popular Times) para evaluar el grado en que dicha fuente online es útil para comprender, analizar y predecir mejor las conductas de los consumidores turísticos.

Siguiendo seis hipótesis sobre el comportamiento de los clientes turístico visitante, se desarrolló una herramienta de software especialmente diseñada, probada con anterioridad y posteriormente se utilizó para obtener una muestra de datos a gran escala de 20.000 períodos de tiempo para 198 restaurantes. Se utilizaron tanto la regresión lineal bi-variante como los análisis de correlación para probar las hipótesis.

Se apoya la hipótesis que incluyen la cantidad de comentarios de los clientes, los efectos de tiempo, el número de imágenes y la información del segmento de precios sobre la cantidad de turistas que visitan un restaurante determinado en promedio. Además, se encuentra una relación positiva con el tiempo de duración promedio. Los hallazgos demuestran que los datos proporcionados a través de Google Popular Times coinciden en alto grado con las suposiciones teóricas y lógicas. Por lo tanto, la fuente de datos es potencialmente eficaz para proporcionar información valiosa a los interesados (por ejemplo, investigadores, administradores, turistas).

Este ensayo es el primero que demuestra conceptual y empíricamente la practicidad y el valor de Google Popular Times para entender, analizar y predecir mejor el comportamiento del consumidor turístico. Por lo tanto, el valor es proporcionado por el potencial de este enfoque para ofrecer datos de comportamiento basados en la comprensión. Es importante señalar que hasta ahora ese enfoque para reunir y analizar ese volumen de datos reales sobre los clientes se consideraba menos práctico en términos de tiempo y gastos.

Although software systems used to automate business processes have been becoming rather advanced, the existing practice of developing and modifying graphical process models in those software systems is still primitive: users have to manually add, change, or delete each node and arc piece by piece. Since such manual operations are typically tedious, time‐consuming, and prone to errors, it is desirable to develop an alternative approach. This paper aims to address this issue.

In this paper, a novel, human‐understandable process manipulation language (PML) for specifying operations (e.g. insertion, deletion, merging, and split) on process models is developed. A prototype system to demonstrate PML is also developed.

The paper finds that manipulation operations on process models can be standardized and, thus, can be facilitated and automated through using a structured language like PML.

PML can improve manipulation operations on process models over the existing manual approach in two aspects: first, using PML, users only need to specify what operations are to be performed on process models, and then a computer carries out specified operations as well as performs other routine operations (e.g. generating nodes and arcs). This feature minimizes user effort to deal with low‐level details on nodes and arcs. Second, using PML, users can systematically specify operations on process models, thus reducing arbitrary operations and problems in process models.