Abstract
In this paper we present an interface based on a recent generative model, the counting grid, here re-introduced in its basic version and largely revised to allow it to deal with large corpora. We show that it is possible to visualize thousands of high order word cooccurrence patterns by only viewing for a few minutes a new embedding we propose for text visualization, browsing and search purposes. We performed preliminary experiments with user tasks such as word spotting, rapid content search and collateral information acquisition
Original language | English |
---|---|
Title of host publication | Proceedings of the ACM SIGKDD 2014 Fulll-day Workshop on Interactive Data Exploration and Analytics |
Editors | unknown |
Place of Publication | u |
Publisher | ACM |
Pages | 107-116 |
Edition | Peer Reviewed |
Publication status | Published - 2014 |
Event | Workshop on Interactive Data Exploration and Analytics (IDEA14) - San Francisco, California, USA Duration: 1 Jan 2014 → … |
Conference
Conference | Workshop on Interactive Data Exploration and Analytics (IDEA14) |
---|---|
Period | 1/01/14 → … |
Other | Sun Aug 24 00:00:00 AEST 2014 |