In the following, we outline a few use cases which shall emphasize the need for ExpertFinder technology.
Rule-enabled metadata, which be used without high entry barrier, shall facilitate that such scenarios can be realized. In order to guarantee low entry costs we need to reuse, wherever possible, formats, vocabularies and categories which are supported by the tools people use already (vCard, iCal, FOAF, ACM categories are examples of such vocabularies), or which are likely to be widely adopted in the near future.
We will try to identify such re-usable vocabularies for each of these scenarios and within the ExpertFinder initiative devise methods and interfaces in order to combine them.
Use Case 1 (basic): Automatic generation of institution webpages and metadata with default values
Let us assume we want to design a FOAF++ enabled content mangement system to manage our institution's web pages and also the member homepages. All members of the institution are allowed to provide their own metadata as extended foaf files, but, if missing, the institution can also specify some standard policies by means of some default rules. Such rules e.g. allow to aggregate metadata from some 3rd party sources. For instance, imagine, your office mate is too lazy to generate his homepage/metadata-file. No problem his basic data can be aggregated from metadata available the university personnel-database, a default publication list can be generated by the meta-data extracted from DBLP, etc.
Another advantage of this scenario is the following. While an individual person can decide to present the same meta-data provided in her own FOAF-file with his individual look-and feel on her personal homepage, the institution's member page, based on essentially the same data might have a completely different look and feel and extract/merge different data. Persons working at an institution may also inherit the address of the insitution by default, while they may possibly provide their own address and thus override the more general address of the institution.
(Remark: This is a very simple scenario, but very efficient. A main point, compared with current CMS solutions is that the generation of pages such as personal homepages or institution does not rely on a central point of data, but allows for aggregation of diffferenc online sources in a normalized, easy to use fashion, given that publishers follow best practices in maintaining their meta-data.)
Use Case 2 (intermediate/advanced): Human Resource Management
People searching for a job simply publish their CV on their homepages or skilled employees in a company make their skill set and experiences available on the Intranet in an agreed metadata format. Job agencies can deploy agents and crawlers which they feed with their preferred profile and which find automatically suitable persons for a given vacancy. Team building within companies can be empowered by automated processes selecting the right set of employees to successfully complete a given project through semantic matching and rules.
ExpertFinder shall enable such scenarios and dezentralize the process of expert and job finding, as opposed to current central recruitment or corporate portals, just as FOAF itself was aimed to decentralize social networks.
Use Case 3 (intermediate): Review selection
For my new journal/workshop I look for reviewers. I asked the authors during my submission to provide keywords and have found a set of ACM kategories which fit my CfP. Now, using citation indexes, committees of previous conferences etc. published in the agreed metadata format, I can easily define in a decarative rule language (possibly with priorities) what are my selection criteria, or adapt selection criteria of previous workshops, if published by the organizers... A mock-up example using DLV-HEX and OWL has been presented at the Answer Set Programming for the Semantic Web Tutorial at ESWC'06.
The reviewer-examples in Unit 6 of this Tutorial describe the mock-up scenario. Another nice example in Unit 7 is about aligning a meeting schedules via iCal. We used google-calendar with it's iCal export at real-time during the tutorial, was fun! ;-)
Use Case 4 (advanced): Trust and security for privacy-relevant meta-data
It would often be desirable that metadata should be encrypted and that the keys to decrypt it shall be provided on a timely limited basis during a process of rule based negotiation. This is for instance nice if I don't want to disclose my private phone number.
I could have rules to process this: if e.g. the person who wants my phone number calls a service to get my phone number where all persons I know are registered (even the service could simply check the sha1-sum in my foaf-file), that service will send a mail back to that address with the (temproary valid) decryption key. only this way, the person can decode my private number.
Different versions of this scenario, with different credentials, more involved negotiation are imaginable.
Use Case 5 (advanced): "Semantic CORDIS"
The ExpertFinder Initiative is obviously related to community efforts such as the EU's successful CORDIS, and could be positioned as as a semantic enrichment of CORDIS and a refinement to the level of individual researchers, in the direction of decentralization and enrichment of the available information stored in such portals. CORDIS enables institutions to find and contact other institutions for joint research projects. Similar to Use cases 2 and 3 he idea here is to create a kind of decentralized social network for institutions. Such an endavour however combines the reuirements for the previous use cases in a more complex scenario: Trust relevant information about projects/partners shall only be disclosed to trustworthy other parties, reliability of the information provided by different parties needs to be assessed, etc.
Use Case 6 (advanced): Semantic Email Addressing
Semantical Email Addressing (SEA) allows emails to be sent to semantically specified recipients. Instead of subscribing to mailinglists persons can indicate interests semantically. If one wants to send a message to a certain group of people with a common interest, one can send this specifying, e.g., a query "To all people with an interest in FOAF". Currently, a prototype SEA module is built on top of Infomaster.
As mentioned in the paper Semantic Email Addressing: Sending Email to People, not Strings FOAF is a possible ontology that can be used to express interest (using foaf:interest). The use case would extend a SEA module, based on FOAF interest, to involve the use of rules to specify for example connections between interests:
if interested in FOAF then interested in Semantical Technologies
A user that indicates an interest in FOAF will, using the knowledge in the above rule, receive emails that are sent to a semantical address all people interested in Semantical Technologies.