| Links' Seminars and Public Events |   |
| 2020 | |
|---|---|
Fri 16th Oct 11:00 am 12:00 pm | Seminar: Aurélien Lemay Title: ShEx Learning from Typed Graphs Abstract: In knowledge graphs, schemas are becoming a new asset to describe the organization of data. The new world-leading format Shex is becoming a de-facto standard in the industry that allows defining flexible and powerful schemas. In this context, the inference of schemas can become a solution to provide shex expressions that describe already existing data. Typically, the inference starts from untyped graphs. However, these tasks appears to be more complex than expected in general, and is possible only for subclasses of Shex. The inference of schemas from typed graph gives a baseline for those algorithms. Its comprehension allows to understand better the underlying difficulties of the task. It presents already unexpected difficulties. We present an algorithm that infers Shex-defined schemas from fully typed graphs. We also present some encountered difficulties, as well as the limitations of the approach. |
Fri 24th Jul 2:30 pm 4:30 pm | Momar Sakho, PhD defense |
Wed 8th Jan 1:30 pm 3:30 pm | Introduction to argumentation theory  Salle Agora 1, Bâtiment ESPRIT |
| 2019 | |
Thu 19th Dec 11:00 am 1:30 pm | Thèse L. Gallois amphi Bâtiment B Inria |
Fri 13th Dec 11:45 am 1:00 pm | 1. On Parsing Gpath (Jérémy and Antonio) 2. On Nested Regular Expression (Joachim) |
Fri 13th Dec 10:30 am 11:45 am | Repet Lily pour l'équipe "Lille-Salle B31 " |
Tue 24th Sep 10:00 am 11:00 am | Stijn Vansummeren Title: General Dynamic Yannakakis: Conjunctive Queries with Theta Joins Under Updates Abstract: The ability to efficiently analyze changing data is a key requirement of many real-time analytics applications like Stream Processing, Complex Event Recognition, Business Intelligence, and Machine Learning. Traditional approaches to this problem are based either on the materialization of subresults (to avoid their recomputation) or on the recomputation of subresults (to avoid the space overhead of materialization). Both techniques have recently been shown suboptimal: instead of fully materializing results and subresults, one can maintain a data structure that supports efficient maintenance under updates and can quickly enumerate the full query output, as well as the changes produced under single updates. In our work we are concerned with designing a practical family of algorithms for dynamic query evaluation based on this idea, and for queries featuring both equi-joins and inequality joins, as well as certain forms of aggregation. Our main insight is that, for acyclic conjunctive queries, such algorithms can naturally be obtained by modifying Yannakakis' seminal algorithm for processing acyclic joins in the static setting. In this talk I present the main ideas behind this modfication, offset it against the traditional ways of doing incremental view maintenance, and discuss recent extensions such as dealing with general theta-joins. Amphitheater of INRIA Building B. |
