Loading...
Please wait, while we are loading the content...
Similar Documents
Data Integration , Entity Reconciliation and Search in Personal Information Networks
| Content Provider | Semantic Scholar |
|---|---|
| Author | Chakrabarti, Soumen |
| Copyright Year | 2006 |
| Abstract | We present SPIN, a system for Searching Personal Information Networks. SPIN seeks to integrate entities and relations from a wide variety of data sources such as address books, emails, documents on file systems, blogs, messenger sessions, social networks and the Web into the PINDB: a universal graph format that has typed nodes, typed edges, and text associated with the nodes. Using the latest techniques for information extraction and deduplication, SPIN adds “soft edges” to the PINDB that correspond to entity aliases and mentions. SPIN’s search algorithms then use this representation to respond to a hierarchy of useful semi-structured query classes with proximity, aggregation and ranking semantics that improves substantially beyond simple desktop search tools. SPIN is a self-contained Java application and has a modular architecture where other researchers can plug in new approaches to data integration and search. |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | http://www.kdd2006.com/docs/KDD06_Demo_17_SPIN.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
