MapMarker: Extraction of Postal Addresses and Associated Information for General Web Pages

摘要: Address information is essential for people’s daily life. People often need to query addresses of unfamiliar location through Web and then use map services mark down the direction purpose. Although both address are available online, they not well combined. Users usually copy individual from a site paste it another with locate its direction. Such operations have be repeated if multiple listed on single page such as public school list or apartment list. Furthermore, associated has copied included each marker better comprehension. Our research devoted automate above process make combination an easier task users. The main techniques applied here include postal extraction extraction. We apply sequence labeling algorithm based Conditional Random Fields (CRFs) train models Meanwhile, using extracted landmarks, we pattern mining identify boundaries blocks extract address. experimental result shows high F-score at 91% 87% accuracy