Duration
25h Th, 10h Pr, 45h Proj.
Number of credits
Lecturer
Substitute(s)
Language(s) of instruction
English language
Organisation and examination
Teaching in the second semester
Schedule
Units courses prerequisite and corequisite
Prerequisite or corequisite units are presented within each program
Learning unit contents
This course aims to provide an overview of the field of linked semantic data, a key domain bridging several modern IT areas, including artificial intelligence, advanced web initiatives, knowledge graphs, data integration and software engineering.
The course will first cover the conceptual foundations of the representation of semantic knowledge and its use for inferences. Semantic networks and ontologies will be introduced, and historical difficulties in reasoning with semantic networks discussed. Description logics will be presented as a theoretical basis for ontology-based reasoning, with appropriate semantics and inference algorithms.
The course will show how these concepts are reused by the semantic web initiative, the purpose of which is to enrich the web with linked semantic data to make it more usable by machines. The link between description logics and the semantic web ontology language OWL will be developed.
The relationships of linked semantic data with data integration, knowledge graphs, NoSQL graph databases and data lakes will then be explored.
Ontology-based data access will be presented as an approach which allows to query a database directly at the ontology knowledge level, thus supporting intelligent, inference-based interactions with the data.
Finally, the course will illustrate through examples and case studies how semantic data are used in modern industrial areas.
The main topics covered will be:
- Knowledge representation, inferences, and semantic knowledge.
- The Semantic Web resource description framework (the initiative, RDF, RDFS, SPARQL).
- Modern ontologies (types, uses, ontological commitment).
- Description logics (including a summary of useful first order logic concepts) and its role in reasoning with ontologies and semantic knowledge.
- The Web Ontology Language: OWL.
- Ontology Engineering.
- Reasoning with description logics.
- Data integration, graph databases and ontology-based data access.
- Selected business application domains (biomedical informatics, software engine
Learning outcomes of the learning unit
At the end of the course, the students will have gained a broad understanding of the field of semantic data, including its theoretical foundations, application domains and related technologies.
They will have learned the knowledge representation principles and inference mechanisms related to ontologies and the logical formalisms supporting them.
They will also understand the vision, ideas and formal languages of the semantic web, as well as the emerging technologies, leading-edge application areas and some of the open questions related to this field.
Finally, they will have developed a practical experience of the tools and challenges involved in building and using an ontology-based application for a specific domain.
Prerequisite knowledge and skills
There are no prerequisite courses required.
A previous experience of formal logic is useful but not necessary: a summary of the concepts of first order logic necessary to understand the material of the course is provided during the course.
For the project, a programming experience sufficient to understand XML-based standards, the basics of Java and the use of a Java-based API (application programming interface) is expected.
Planned learning activities and teaching methods
Lectures (25 h) will cover the theoretical content of the course.
Practice sessions (10h) will cover:
- exercises on the theoretical foundations;
- an introduction to ontology modeling and to the programming tools needed for the project.
A project will be realized in small groups with the goal to implement an ontology for a specific domain and to access this ontology to retrieve information, using open source tools.
Mode of delivery (face to face, distance learning, hybrid learning)
Note : the information provided here is subject to the condition that, by the time the course starts in February 2021, a normal mode of teaching will be back in place. If Covid restrictions are still present, the mode of delivery will have to be adjusted accordingly.
The theoretical sessions and practice sessions will be delivered face-to-face.
The project will mostly be carried out remotely. A final review with presentation and defense of the results will take place during one of the last sessions of the course.
Organisational adjustments related to the current health context
If, by the time the course takes place (starting in February 2021), the covid situation or the lack of appropriate rooms prevents activities on site,
the following remediations are foreseen depending on the situation :
- courses sessions will be given in remote (using e-campus collaborate or any other suitable application);
- the exam is transformed in an individual oral remote discussion where the main question will be selected out of a predefined list and presented with open books like last year; details are provided in the assessment methods section.
- the project will be assessed by a remote videoconference review.
Recommended or required readings
The reference material for the theory and the practice sessions is provided in the slides of the course.
Due to the diversity of the subject matter there is no global reference textbook. Useful complementary readings will be indicated at the start of each chapter when relevant.
The following sources provide non-required but useful reading on knowledge representation and description logics:
- Chapters 7, 8 and 12 of the book of Russel and Norvig: Artificial Intelligence: A Modern Approach (3rd Edition), Pearson, 2010.
- Chapters 1, 2, 4, 7 and 8 of the book of Baader, Horrocks, Lutz and Sattler: An introduction to description logic, Cambridge University Press, 2017.
The realization of the project will require to consult the online documentation of the semantic web standards and the chosen tools. Appropriate pointers will be provided.
Assessment methods and criteria
Below you will find information on the evaluation methods planned for in-person and remote exams as well as those planned for hybrid sessions. Depending on how the health crisis evolves, the chosen method will be communicated to you no later than one month before the start of the exam session.
Any session :
- In-person
oral exam
- Remote
oral exam
- If evaluation in "hybrid"
preferred in-person
Additional information:
Note : the information provided here takes into account the Covid situation at end of March 2021.
Content of the theory and exercises sessions will be assessed by an oral exam in June, and if required a second-session oral exam in August/September.
The exam will be an individual oral discussion, either in remote or in presential depending on the circumstances. By preference it will be done in presential, if possible. If in remote, the exam will use by preference Collaborate.
In both cases, the exam will be organized in the same manner :
Students will be asked first to present one main question out of a list of high level questions defined in advance. For this question the students will be allowed to use the relevant slides from the theory of the course, and will be asked to explain them.They may also use any notes they may have prepared about these high level questions.
Students will then be asked oral complementary questions testing their understanding and their capability to make links between different chapters of the course. One of these complementary questions will be related to one of the two case studies.
These complementary questions are not limited to the topic of the main question mentionned above but may range over the whole material, and in order to provide answers in timely fashion, students are still required to study the main ideas of the course and to be able to answer them without taking time to research their notes.
Typical examples of complementary questions concern the definition, use, and key characteristics or limitations of the concepts seen in the course and their relationship to other concepts. Students may also be asked to resolve or explain an example covered in the theory section of the course, and when doing so, are also expected to be able to talk about the concepts illustrated by that example.
The required material for the exam covers chapters 1 to 11 ot the theory and the two case studies.
To take into account the Covid situation:
- the material from the practice sessions is suppressed from the material for the exam (but still usefull for the project).
- some specific sections or slides have been excluded from the theory chapters. They are explicitely identified at the start of each chapter (starting with chapter 6).
The project will be assessed by a remote videoconference review, using Collaborate.
The project results will be assessed from the resulting implementation, and a final defense, including a presentation and a demonstration.
Timely submission of project results is mandatory for presenting the exam!
Given the circumstances and the suppression of the practice material from the oral exam, the ponderation of the oral exam is reduced in the final note, as last year :
Oral exam : 50% of the final note;
Project : 50% of the final note.
Grades for the project are normally assigned to the whole group. However, in some cases (e.g. when there is evidence that a member of a group has not participated enough in the project), the grade may be assigned individually, reflecting the personal involvement of each member of a group.
Students who must represent the course in the August/September session may opt either to keep the grade obtained for the project in June or to improve their project. This improvement will have to be carried out individually if no member of the same project group is in the same situation. There is no support guaranteed during the summer for project improvement.
Work placement(s)
Non applicable.
Organizational remarks
The course takes place in second semester; the exact start date and location will be posted on the course web page.
The project will begin a few weeks after the start of the course (to allow the students to acquire the necessary knowledge); its exact start date will be posted on the course web page. The project results will be presented during one of the last sessions of the course.
Contacts
Lecturer: Jean-Louis Binot (jean-louis.binot@uliege.be).