Titlebook: Web and Big Data; 5th International Jo Leong Hou U,Marc Spaniol,Junying Chen Conference proceedings 2021 Springer Nature Switzerland AG 202

終止

被詛咒的人

BOUNCE: An Efficient Selective Enumeration Approach for Nested Named Entity Recognitionwhich acts as a filtering step before enumeration. We provide a detailed analysis on the time complexity of the existing nested NER techniques and conduct extensive experiments on two datasets. The results demonstrate that BOUNCE outperforms various nested NER techniques and achieves higher efficien

confederacy

喃喃訴苦

fibroblast

強制性

Multi-space Knowledge Enhanced Question Answering over Knowledge Graphcially designed attention mechanism is used to calculate the score of candidate answers dynamically. The experiment conducted on the WebQuestions dataset shows that compared with other state-of-art QA systems, our method can effectively improve the accuracy.

floodgate

鞏固

AIR Cache: A Variable-Size Block Cache Based on Fine-Grained Management MethodFine-Grained Recorder (FGR) where FGR is dedicated to record the data popularity using fine-grained data sections, MGW writes data at different granularity, and MGE is responsible for evicting the data at dynamic granularity. Our experiments with real-world traces demonstrate that AIR cache can incr

falsehood

AIR Cache: A Variable-Size Block Cache Based on Fine-Grained Management MethodFine-Grained Recorder (FGR) where FGR is dedicated to record the data popularity using fine-grained data sections, MGW writes data at different granularity, and MGE is responsible for evicting the data at dynamic granularity. Our experiments with real-world traces demonstrate that AIR cache can incr

Nonflammable

SardineDB: A Distributed Database on the Edge of the Networkd random read performance of SardineDB have great advantages compared with existing distributed databases on the edge. As a result, SardineDB is very suitable for edge because it has high write performance, low GC burden and low write amplification.