semweb2slide86
From CommerceNet Wiki
| Prev Slide | Index | Next Slide |
Our final example is an e-life scenario: planning a dinner for tonight in Pittsburgh.
The task breaks down into a series of subtasks - determining who is in town that I might like to spend some time with, if they available and interested in meeting for dinner, finding a restaurant, booking reservations, making arrangements, and coordinating any last minute changes in plans.
These are precisely the types of tasks the Web excels at.
Let’s start with who’s in town that I might like to meet. That list would include people I know who live in Pittsburgh, as well as conference attendees with whom I have something in common – perhaps authors of papers I’ve liked, or colleagues who liked the same papers I did, or those who fund research in my area.
I might start by asking my Personal Agent to compile a list of the people I know who live in Pittsburgh, using contacts from my address book and social networking sites.
To compile a list of conference attendees, someone would have to create a service that scraped the conference site for speakers, session chairs and the like, and published the results as a microformatted list. In the future, conference organizers might routinely publish such structured lists of registered attendees. Someone else might build a collaborative filtering service for CiteSeer, where users share ratings of papers. Given such services, a simple agent could correlate the list of conference attendees with the authors of papers I liked to suggest possible dinner guests.
Given a list of potential dinner guests, an agent could use email or a service like Evite to see who’s available. An incremental service, built on Evite, might offer responders the opportunity to fill out a micro-form, indicating preferences regarding cuisine, location, etc.
Next up is selecting a place to eat. There are numerous sites on the Web that rate and review restaurants. A helpful knowledge service could use these existing sites, together with guests’ indicated preferences, to recommend restaurants. Another knowledge service could then check availability at Open Table, book the reservation, and notify the guests.
Each of these knowledge services is helpful in its own right, so it’s a good bet that someone will create services like them. Once they exist, it won’t take long before someone drops in a constraint-based planner and offers a completely automated dinner planning service.
