Article Preview
TopIntroduction
With today's technology any type of information accessibility has greatly increased and this is also true for our domain of study: Call For Papers (CFP). It has never been so easy to find information on conference usually via numerous websites or simply by email. However, this profusion of information has also its downside. Handling CFP can be seen as a really time consuming task. Regarding the emails, many CFP are received every day and most of them will only be skimmed through or forgotten by lack of time or attention. As for the web, it is possible to find many websites where you can browse a wide range of CFPs such as WikiCFP (2014) or ConferencePartner (2014), or others bound to different institutions like IEEE (2014) or ACM (2012). Nevertheless these sites have several weak points. They do not allow the user to use CFPs obtained by external sources (e.g. Email) ; thus a new research have to be done on a totally different database, sometimes with less information you could actually find in your CFP email. Moreover, conferences research parameters can be really basic and not handle a request more complex than searching different strings of characters (e.g. conference about X topic in Y country with a Z deadline). Those are the issues we propose to handle with our system: CFP Manager.
In one of our previous work (Issertial & Tsuji, 2011), we proposed the concept of a text mining system able to extract relevant information from a group of CFP. In another one concerning visualized comparison of CFP datasets (Issertial, Saga, & Tsuji, 2012), we introduced the idea of extracted data enrichment via the utilization of ontology, another one (Issertial & Tsuji, 2013) focused on the query and interface system along with the different kind of output that can be obtained from the set of relevant data previously collected. This paper proposes to focus on the case of a final user who’s looking for information about conferences suited to him and how our proposed system will ease this process. We will do so by reviewing current web-services proposing this functionality and comparing their theoretical results with our proposed system on different queries. This will be introduced along with the description of our system and numerical experimentations proving its good working. CFP will be subject to rule based text mining algorithms in order to extract relevant information. These very data will be enriched via ontology models with concepts close to the ones found in the CFP. Finally, using the previously collected data, we form the base of a user interface system with an intuitive query system allowing the user to perform complex queries.