One of our internal teams (LanguageWare-- they provide the spellchecker and some stuff for Live Text, and other stuff) is going to betaking a full week out to focus entirely on "entity extractiondevelopment" which is geekspeak for "recognizing what a word is"-- "Hannover" being a city OR a development project, things likethat. So if you have some entities which you really would like tosee included please let us know.
By entities we mean any named thing in text, like projects,technologies, people, times, companies, organizations, places,addresses, legal jargon, ... etc.. this is your chance to suggestsomething a bit more off the wall, or even very mainstream and domainspecific.
Here are some things I know Samantha would want to be automatically identified:
- Dates/Times and roles (due date, sent date, event date, etc.)
- Relative Dates/Times: "today", "tomorrow", "noon next Tuesday", etc.
- Scheduled calendar events (meetings, conferences, etc.)
- Aggregate names, from lists supplied by other components, like Projects, Activities, etc.
- Information place names, like QuickPlace, Quickr
- Community names from Connections, etc.
- Geographical place names
- Patent numbers and patent publication numbers
- Publication names (print or electronic)
- Publication article titles