Reinforcement Learning for Knowledge Graph Reasoning | Victoria Lin | Knowledge Connexions 2020 скачать в хорошем качестве

Reinforcement Learning for Knowledge Graph Reasoning | Victoria Lin | Knowledge Connexions 2020

Large-scale knowledge graphs (KGs) support a variety of downstream NLP applications such as semantic search and chatbots. Answering arbitrary user questions over KGs often requires reasoning over multiple inter-related facts. In a common setup where the topic entity and the target relation are known, the problem can be formulated as a partially observable Markov decision process (POMDP), where a policy-based agent sequentially extends its inference path from the topic entity until it reaches a target. However, in an incomplete KG environment, the agent receives low-quality rewards corrupted by false negatives in the training data, which harms generalization at test time. Furthermore, since no golden action sequence is used for training, the agent can be misled by spurious search trajectories that incidentally lead to the correct answer. In this work, Victoria Lin proposes two modeling advances to address both issues: (1) Reducing the impact of false negative supervision by adopting a pretrained one-hop embedding model to estimate the reward of unobserved facts; (2) Countering the sensitivity to spurious paths of on-policy RL by forcing the agent to explore a diverse set of paths using randomly generated edge masks. this approach significantly improves over strong path-based KGQA baselines and is comparable or better than embedding-based models on several benchmark datasets. -- Victoria Lin: Senior Research Scientist, Salesforce Research Victoria's primary research interests center around the intersection of natural language and structured data, which include jointly reasoning over and translation between the two modalities. Lately she has been focusing on natural language based database querying, which requires the understanding of user utterances grounded in complex database structures. Prior to joining Salesforce in 2017, she was a graduate student in computer science at the University of Washington, where she introduced NL2Bash, a popular benchmark for building natural language shells. -- --- Welcome to Connected Data London's #ThrowbackThursday Every Thursday at 3pm GMT, we are releasing gems from our vault on #YouTube Tune in and learn from leaders and innovators; subscribe to our channel and watch premieres as they are released! #KnowledgeGraph #Ontology #DataModeling #AI #KnowledgeRepresentation #EmergingTech #SemanticWeb #SemTech