Companies in China Company A Chinese company Company

“理解”“名词短语” • 通常来说，一个名词短语指示一个对象 • • Companies in China -> 一些Company A Chinese company -> 一个Company Their company -> 特定的Company Largest fintech company in China -> 蚂蚁金服

“理解”“名词短语” “You shall know a word by the company it keeps” ——Firth (1957) Question: what is Obama’s citizenship? Query parsing: (Obama, Citizenship, ? ) Identify and infer over relevant subgraphs: (Obama, Born. In, Hawaii) (Hawaii, Part. Of, USA) correlating semantically relevant relations: Born. In~Citizenship Answer: USA MSR-DL-Summer School

QALD上的“复杂名词短语” • 常见模式 NP 1 wh-word VB (IN) NP 2 NP 1 (VBN) IN NP 2 NP 1 of NP 2 N 1 N 2 • 常见结构 NP ? NP NP ? ? total population of Melbourne members of the Star Alliance films starring Clint Eastwood agencies in the Maldives the owner of Facebook Father of Singapore 附着动作 • 求序数 the first with highest • 计总数 more than 3 the number of • ……

经典(结构化)知识库问答方法 • CCG Parsing (Yoav Artzi) • Parsing on QA-Pairs (Percy Liang) • Staged Query Graph (Scott Yih) • Subgraph Matching (Lei Zou)

CCG Parsing (Yoav Artzi) • Scaling Semantic Parsers with On-the-Fly Ontology Matching • 理解 • 49 domain independent lexical items • 56 underspecified lexical categories • 求解 • Structure Match • Constant Matches

Parsing on QA-Pairs (Percy Liang) • Unary Binary Entity Join Property(Entity, ⋅) Intersection Aggregation Count(unary)

Staged Query Graph (Scott Yih) • Semantic Parsing via Staged Query Graph Generation: Question Answering with Knowledge Base • “理解”/求解 “Who first voiced Meg on Family Guy?

Subgraph Matching (Lei Zou) • Natural Language Question Answering over RDF —A Graph Data Driven Approach • “理解”/求解

“理解”与“求解”的目标差 异 • 理解“难”和求解“难”不是同一个“难” How old is Michael Jordan? • 结构：Age(Michael Jordan) • 求解： Give me all writers that won the Nobel Prize in literature. • 结构：&^$%$^@!#*^@@! • 求解：Wikipedia: List_of_Nobel_laureates_in_Literature ? x ? p dbpedia_categories: Nobel_Prize_in_Literature ? x a dbo: Writer

Thanks for listening • Q&A