麻豆社

Argomenc dl curs
The Data Preparation and Integration module addresses a variety of problems related to the integration of heterogenous data sources. It overviews the main issues in data integration, notably handling different forms of heterogeneity, and presents the general architecture of data integration systems. Foundational techniques for data integration are covered, such as data matching, schema matching and mapping, and query processing in data integration. A specific data integration approach relying on the technology of Virtual Knowledge Graphs and semantic mappings is presented in detail. The integration both of relational data sources, and of other types of data sources accessed by relying on data federation technology are considered. By attending the course, students will learn how to design and build a comprehensive data integration solution, possibly exploiting existing data access and data federation technologies. The Data Profiling module considers a variety of problems related to the profiling of relational and non-relational data. It first overviews techniques for the exploratory analysis of relational databases, which are also adopted in preparatory activities of data cleansing and integration, and mining. Widely used statistical measures, such as the Pearson and Spearman correlation coefficients, are discussed and applied to real word examples in the first part of the course. Special attention will be also paid to the Phi-K correlation coefficient, which has been recently proposed to overcome well known limitations of the former two, such as the possibility to analyse correlations between numerical, categorical, interval and ordinal variables. Algorithms for discovering patterns and relational dependencies in tabular data will be presented in detail. In particular, the main part of the course will go through the formal specifications of uniqueness, functional and inclusion dependencies in relational data, with a special focus on state-of-the-art algorithms and their complexity. If time permits, the last part of the course will discuss the application of the introduced profiling techniques to the relational query optimization problem, and to a high level introduction of recent profiling methods that can be applied to RDF data.

Modalit茅 de ensegnament
Frontal lectures, exercises, and labs.

Obietifs formatifs
The course belongs to the type "caratterizzanti 鈥� discipline informatiche" in the curriculum 鈥淒ata Analysis鈥�. The Data Integration module addresses a variety of problems related to the integration of heterogenous data sources. It overviews the main issues in data integration, notably handling different forms of heterogeneity, and presents the general architecture of data integration systems. Foundational techniques for data integration are covered, such as data matching, schema matching and mapping, and query processing in data integration. A specific data integration approach relying on the technology of Virtual Knowledge Graphs and semantic mappings is presented in detail. The integration both of relational data sources, and of other types of data sources accessed by relying on data federation technology are considered. By attending the course, students will learn how to design and build a comprehensive data integration solution, possibly exploiting existing data access and data federation technologies. The Data Profiling module considers a variety of problems related to the profiling of relational and non-relational data. It first overviews techniques for the exploratory analysis of relational databases, which are also adopted in preparatory activities of data cleansing and integration, and mining. Widely used statistical measures, such as the Pearson and Spearman correlation coefficients, are discussed and applied to real word examples in the first part of the course. Special attention will be also paid to the Phi-K correlation coefficient, which has been recently proposed to overcome well known limitations of the former two, such as the possibility to analyse correlations between numerical, categorical, interval and ordinal variables. Algorithms for discovering patterns and relational dependencies in tabular data will be presented in detail. In particular, the main part of the course will go through the formal specifications of uniqueness, functional and inclusion dependencies in relational data, with a special focus on state-of-the-art algorithms and their complexity. If time permits, the last part of the course will discuss the application of the introduced profiling techniques to the relational query optimization problem, and to high level introduction of recent profiling methods that can be applied to RDF data. Knowledge and understanding: 鈥� D1.1 - Knowledge of the key concepts and technologies of data science disciplines 鈥� D1.2 - Understanding of the skills, tools and techniques required for an effective use of data science 鈥� D1.6 - Knowledge of the principles and methods of data curation Applying knowledge and understanding: 鈥� D2.1 - Practical application and evaluation of tools and techniques in the field of data science 鈥� 2.5 - Ability to apply, evaluate and develop methods and tools for the integration, cleaning, and quality of data 鈥� 2.10 - Application of languages, tools, and methods for the design of information systems and their corresponding software applications for data, process, and organization management Making judgments 鈥� D3.2 - Ability to autonomously select the documentation (in the form of books, web, magazines, etc.) needed to keep up to date in a given sector Communication skills 鈥� D4.1 - Ability to use English at an advanced level with particular reference to disciplinary terminology 鈥� D4.3 - Ability to structure and draft scientific and technical documentation Learning skills 鈥� D5.2 - Ability to autonomously keep oneself up to date with the developments of the most important areas of data science

Sort de ejam
Oral exam and project work. The mark for each part of the exam is 18-30, or insufficient. The oral exam covers Module 1 鈥淒ata Integration鈥� and Module 2 鈥淒ata Profiling鈥�, and comprises verification questions, and open questions to test knowledge application skills. It counts for 50% of the total mark. The project consists of two parts. Part 1 covers Module 1 鈥淒ata Preparation and Integration鈥� and verifies whether the student is able to apply advanced data integration techniques and technologies taught or presented in the course to solve a concrete integration problem. It is assessed through a final presentation, a demo, and a project report and can be carried out either individually or in a group of 2 students. It is discussed during the oral exam, and it counts for 25% of the total mark. Part 2 covers Module 2 鈥淒ata Profiling鈥� and verifies whether the student is able to apply data profiling techniques to a concrete use case. It is assessed through a final presentation and a project report and can be carried out either individually or in a group of 2 students. It is discussed during the oral exam, and it counts for 25% of the total mark.

Criters de valutazion
The final mark is computed as the weighted average of the oral exam, Part 1 of the project, and Part 2 of the project. The exam is considered passed when all three marks are valid, i.e., in the range 18-30. Otherwise, the individual valid marks (if any) are kept for all 3 regular exam sessions, until also all other parts are completed with a valid mark. After the 3 regular exam sessions, all marks become invalid. Relevant for the oral exam: clarity of answers; ability to recall principles and methods, and deep understanding about the course topics presented in the lectures; skills in applying knowledge to solve exercises about the course topics; skills in critical thinking. Relevant for the project: skill in applying knowledge in a practical setting; ability to summarize in own words; ability to develop correct solutions for complex problems; ability to write a quality report; ability in presentation; ability to work in teams. Non-attending students have the same evaluation criteria and requirements for passing the exam as attending students.

Bibliografia obligatora

A. Doan, A. Halevy & Z. Ives (2012). Data Integration. Morgan Kaufmann. (ST270 D631)

Z. Abedjan, L. Golab, F. Naumann, & T. Papenbrock (2018). Data Profiling. Synthesis Lectures on Data Management, 10(4), 1-154.

Z. Zhang, R. Zhang (2008). Multimedia Data Mining: A Systematic Introduction to Concepts and Theory. CRC Press LLC.

Additional material (slides, notes of the lecturers) will be made available before each lesson.

 

Subject Librarian: David Gebhardi, David.Gebhardi@unibz.it

Bibliografia aconsieda

R. C. Gonzalez, & R. E. Woods (2007). Image processing. Digital image processing, 2, 1.

R. O. Duda, P. E. Hart & D. G. Stork (2012). Pattern classification. John Wiley & Sons. 

Depl霉 informazions
Software used: Ontop system for deploying and querying Virtual Knowledge Graphs, developed by the In2Data research group at the Faculty of Computer Science. Relational DBMS, such as PostgreSQL. Data federation tools such as Denodo, Dremio, Teiid. Python3.5 with pandas, scikit-learn, scikit-image.