Service Description and Discovery Scenario - SOA4All - UIBK

From FutureInternetWiki

Future Internet Scenario: Microservices

Developed economies increasingly concentrate on the service industry, in contrast to manufacturing and agriculture. Both the GDP output and labor force are mostly in services. While historically, the service sector tended to be less affected by international (and inter-regional) competition than manufacturing, dramatic cost reductions and speed and reliability improvements in the transportation of people and the communication of information lead to increasing the range of service provision. [CIA World Factbook, Wikipedia] Information technologies are on their way to helping significantly increase the availability and competition in services. Good examples of this phenomenon are online retailers (esp. mass services such as Amazon) and personal commerce (marketplaces such as eBay), with direct benefits of greater availability and lower prices of products. We sketch how technology can extend these benefits past sale and resale services, and enable specialized micro-services that thrive economically thanks to good matching with customers. The project SOA4All is currently researching parts of the necessary technologies.

What if we can describe services in such a way that it becomes easy to discover and use then over the Web? In comparison, the structure of product descriptions has been well explored in decades of cataloguing and advertising in support of manufacturing: full-text search combined with some categorization and recommendation systems, together with detailed description of delivery, are apparently sufficient for large-scale sale operations (both retail and personal commerce). First and foremost, the technology must focus on the ease of finding services. Here are some scenarios:

1. When organizing a wedding, we want a good hairdresser for the bride; they need to be able to come to our place on a given date. The system would rank the available hairdressers by their reputation, location and possibly other criteria (such as prices), and the user would then select one, for instance by also looking at samples of work. The system would make a reservation with the selected hairdresser.
2. Continuing with the wedding scenario, an eccentric millionaire father of the bride wants the best iced cake for the wedding, price and location is no object. For cake bakers that cannot arrange delivery to the wedding location, try to find a delivery service that will care for the cake properly. The user (the millionaire father) would likely select a baker mostly by their reputation and by special possible features (skillful application of gold-leaf, original structure etc.); even be a local one-person operation would have a chance of being selected and the system would initiate the order.
3. A guest at the wedding knows the bride really appreciates ornate drapery with lily patterns; if eBay and other services can find no suitable ready-made products, the guest may look for specialist drapers and make an order for a unique hand-made set as the wedding present; a skilled hobbyist draper with good reputation (and with lower prices than a specialist shop because the hobbyist has day-job income) could be selected if the guest is not as wealthy as the bride’s father. The system should enable the hobbyist to easily set up an order/inquiry form.
4. And to get back to hairdressers, a hip young lady traveling in an unfamiliar city has just met someone special, managed to arrange a date for tonight. Among the urgent preparations is a good hairdo: she can use her smart-phone to find a close-by hairdresser with a suitable style, plus directions on how to get there.
5. Finally, that same hip young lady is keen on wearing hats; she can set up her smart-phone to monitor for the presence of specialty hat stores in her vicinity as she travels.

Because location, time and other context is crucial with services (and much less so with product sales), a service description and discovery technology that would enable such scenarios (hair-dressing, cake-baking, drapery-weaving, façade-painting, text proof-reading etc.) will naturally combine with mobile technologies. On the client’s side, a mobile device will be used for discovering service providers (and for direct contact, if need be); on the provider’s side, good integration with the calendar and the current location of the provider would enable more precise discovery and quick response (“I’m in the neighborhood, I can meet you in 20 minutes”). As shown, the technology encompasses all kinds of services. We have especially focused on low-scale services in the long tail of service economy, similarly to how eBay enables small stores to find more customers, and individuals to sell what they do not need. With such technology, hobbyists would be much more likely to find opportunities to make money while doing their hobby. Small shops with specialist expertise would easier find customers. Like eBay enabled personal commerce, SOA4All might enable personal services.

Needed technologies (research directions):

1. Service description that makes it easy to describe a real-world service in enough detail for useful discovery.
2. Discovery based on functionality, location, access range, pricing, and importantly feedback.
3. Trusted feedback and reputation.
4. Computer-assisted ordering (invocation) of a service.