wiki_research

personal research wiki
git clone https://a3nm.net/git/wiki_research/
Log | Files | Refs

commit 9568f40f1b591fb0739defc5dfe10023229e1ed8
parent 33821430c4dd2ca148491c1a121b6c9f14f16b7c
Author: Antoine Amarilli <a3nm@a3nm.net>
Date:   Wed, 25 Jun 2025 18:02:33 +0200

commit with codex

Diffstat:
aggregated_deletion_propagation | 9+++++++++
causal_responsibility | 10++++++++++
database_modification | 2++
deletion_propagation | 2+-
deletion_propagation_with_source_side_effects | 2++
matrix_types | 1+
np_complete | 2+-
probabilistic_databases | 2++
set_cover | 2++
tuple_deletion | 2+-
vertex_cover | 2+-
11 files changed, 32 insertions(+), 4 deletions(-)

diff --git a/aggregated_deletion_propagation b/aggregated_deletion_propagation @@ -0,0 +1,9 @@ +# Aggregated deletion propagation + +[Deletion_propagation] where you want to delete some number of [query_answers] (but it is not specified which ones) + +Studied in [hu2020aggregated] + +Up: [deletion_propagation] + +Aliases: ADP-SS, ADP SS diff --git a/causal_responsibility b/causal_responsibility @@ -0,0 +1,10 @@ +# Causal responsibility + +For a [Boolean_query] Q and [fact] F in an input [relational_database] D, we say that a tuple has "causal responsibility k" if D satisfies Q but there is a set Γ of at most k facts such that D\Γ satisfies Q but D\(Γ cup {F}) does not satisfy Q +- i.e., removing Γ makes F "counterfactual", i.e., it becomes critical for the query to be true + +- [meliou2010complexity]: [dichotomy] on [SJFCQs] in [set_semantics] +- [makhija2023unified]: [dichotomy] on [SJFCQs] in [bag_semantics] + - goes via [MILP] + +Up: [database_theory] diff --git a/database_modification b/database_modification @@ -6,3 +6,5 @@ See also: [graph_modification] Up: [RDM] + +Aliases: database modifications diff --git a/deletion_propagation b/deletion_propagation @@ -7,7 +7,7 @@ Find the minimal set of [facts] to delete from a [database] so that the query re - special case (for [Boolean_queries]): [query_resilience] - [deletion_propagation_with_view_side_effects] - [aggregated_deletion_propagation] - - [smallest_witness_problem] + - [smallest_witness_problem]: do as many [tuple_deletions] as possible without changing the [query_answers] A special case of [RDM] where the only [database_modifications] permitted are [tuple_deletions] diff --git a/deletion_propagation_with_source_side_effects b/deletion_propagation_with_source_side_effects @@ -7,3 +7,5 @@ Special case (for [Boolean_queries]): [query_resilience] Up: [deletion_propagation] See also: [deletion_propagation_with_view_side_effects] + +Aliases: DPSS, DP_SS diff --git a/matrix_types b/matrix_types @@ -19,5 +19,6 @@ - [strongly_unimodular] matrix - [pet_matrix] - [matrix_symmetric] +- [matrix_balanced] Up: [matrix] diff --git a/np_complete b/np_complete @@ -26,4 +26,4 @@ Up: [completeness] for [nptime] See also: [intractability], [np_hard], [conp_complete] -Aliases: NP-completeness, NP_completeness +Aliases: NP-completeness, NP_completeness, NPc diff --git a/probabilistic_databases b/probabilistic_databases @@ -28,3 +28,5 @@ See also: [relational_algebra], [relational_calculus] Up: [database_theory] + +Aliases: probabilistic database diff --git a/set_cover b/set_cover @@ -7,3 +7,5 @@ Given a universe [set], and a collection of [subsets] whose [union] is the unive See also: [hitting_set], [vertex_cover], [edge_cover] Up: [set] + +Aliases: vertex cover hypergraphs diff --git a/tuple_deletion b/tuple_deletion @@ -4,6 +4,6 @@ The [deletion] of a [tuple] from a [database_instance] Up: [database_modification] -Aliases: tuple delete +Aliases: tuple delete, tuple deletions See also: [tuple_insertion], [edge_deletion] diff --git a/vertex_cover b/vertex_cover @@ -1,6 +1,6 @@ # Vertex cover -A vertex cover in a [graph] is a subset of vertices such that every edge is incident to at least one vertex of the subset. +A vertex cover in a [graph] is a [subset] of [vertices] such that every [edge] is incident to at least one [vertex] of the [subset]. [linear_relaxation]: [fractional_vertex_cover] - and [dual]: [fractional_edge_packing]