The YAGO Knowledge Base CC-BY Fabian M. Suchanek >lm&kbs >kbs&dbs
Language Models and Knowledge Bases •   designed to generalize (applies to machine learning models in general) •   designed to memorize (applies to structured data in general) 2 Language Models Knowledge Bases
Language Models and Knowledge Bases •   designed to generalize (applies to machine learning models in general) •   designed to memorize (applies to structured data in general) 3 Language Models Knowledge Bases You give the blue dots it generalizes to a line it stores the two dots
Language Models and Knowledge Bases •   designed to generalize probabilistic opaque costly inference •  creative/ flexible =>   ideal for natural language processing  (applies to machine learning models in general) •   designed to memorize deterministic transparent cheap inference •  “dumb” / rigid => ideal for storing crisp data (applies to structured data in general) 4 Language Models Knowledge Bases
Databases and Knowledge Bases •   entities are stored in a table and have exactly the attributes of table fast on entity‐centric queries has reasoning via constraints •  not made for taxonomies =>   ideal for complete data and “flat” data •    entities are stored in a graph and can have (nearly) any attributes •  need lots of joins for such queries •  have good reasoning support •  have a taxonomy, inherited properties =>   ideal for incomplete data and data with taxonomic hierarchy 5 Databases Knowledge Bases 1 2 1     2   /    / 5
6 Knowledge Base: Inheritance Relations (properties, attributes) are declared for a class, and inherited by the subclasses. Human Entity type subclass subclass Singer Elvis_Presley_(singer) - label - URL - image - birth date - birth place - spouse - song - awards
7 Knowledge Base: Terminology Knowledge Base (KB, also: Knowledge graph, KG) is a directed labeled graph, where the nodes are entities and the edges are relations between these entities.  Human Entity type subclass subclass Song CreativeWork subclass subclass type sang Taxonomy of classes Instances with facts Singer Elvis_Presley_(singer) Blue_Suede_Shoes_(song) Labels “Elvis” “Blue Suede Shoes”  “Синие замшевые туфли”   label label - label - URL - image Relation definitions
8 Knowledge Base: Terminology Knowledge Base (KB, also: Knowledge graph, KG) is a directed labeled graph, where the nodes are entities and the edges are relations between these entities.  Human Entity type subclass subclass Song CreativeWork subclass subclass type sang Taxonomy of classes Instances with facts Singer Elvis_Presley_(singer) Blue_Suede_Shoes_(song) Labels “Elvis” label label Constraints: Human⊓  CreativeWork ≡ ⊥   “Blue Suede Shoes”  “Синие замшевые туфли”   - label - URL - image Relation definitions
9 Knowledge Base: Terminology Knowledge Base (KB, also: Knowledge graph, KG) is a directed labeled graph. Human Entity type subclass subclass Song CreativeWork subclass subclass type sang Taxonomy of classes Instances with facts Singer Elvis_Presley_(singer) Blue_Suede_Shoes_(song) Labels “Elvis” label label Constraints: Human⊓  CreativeWork ≡ ⊥   “Blue Suede Shoes”  “Синие замшевые туфли”   - label - URL - image Relation definitions Schema/ontology
10 [Noy &al: Industry-Scale Knowledge Graphs — Lessons and Challenges, CACM 2019] = where I gave a talk [Weikum, Dong, Razniewski, Suchanek: Machine Knowledge, Found. & Trends in Databases, 2021] Knowledge Bases in Industry
11 [Weikum, Dong, Razniewski, Suchanek: Machine Knowledge, Found. & Trends in Databases, 2021] Knowledge Bases in Industry: Construction