Other Calendars
AUA Internal Calendar | Academic Calendar | Admissions Calendar| Open Education Calendar | Library Calendar| Center for Student Success Calendar
AUA Public Events
- This event has passed.
Recommendation Systems and Meta-Analysis
September 29, 2017 @ 12:30 pm - 1:30 pm +04
About the Event:
It is now well established that recommendation systems, based on a collaborative filtering approach, bring real value to users.
The first part will deal with the application of a recommendation system in an OLAP context (On-Line Analytical Processing, for the analysis of data cubes). In particular, a system recommending sequences of OLAP queries, combining extraction of association rules from logs and similarity measures, will be detailed. This part will conclude with the presentation of a Java prototype, called Falseto for “Former Analytical Sessions for Tedious Olap.” This tool is used to assist a user in the composition of queries and sessions by using holistic functions such as log summaries (hierarchized by a clustering technique), navigation operators, and recommendations.
The second part will focus on the analysis of texts on cultural heritage data. The joint use of frequent patterns and topic models, represented in the form of a graph, will be discussed. There will also be guidelines for the recommendation of words/objects of cultural heritage, based on this graph.
The third part will deal with a work in progress on a workflow recommendation system (i.e. a chain of data-mining operators) for the analysis of data. The particularity of this recommendation system is to only exploit the context in which a workflow is executed: the dataset to be analyzed and the performance indicators desired for this analysis.
About the Speaker:
Dr. Julien Aligon currently Maitre de Conférences (Associate Professor) at the University of Toulouse 1 Capitole and is a member of the IRIT lab (SIG team). He is also an external member of the LI of the University of Tours (France) where he received his PhD in 2013.
His research topic is mainly focused on the field of database analysis, in particular: recommender and personalization systems, similarity measures, log mining, data warehouses, OLAP. He had also the opportunity to work on use cases of data mining algorithms (especially frequent patterns) applied for cultural heritage and logs of car configurations. He is also interested in hadoop systems.
