|Title||Creating dynamic personalized RSS summaries|
|Publication Type||Conference Paper|
|Year of Publication||2008|
|Authors||Bouras, C, Poulopoulos, V, Tsogkas, V|
|Conference Name||8th Industrial Conference on Data Mining – ICDM 2008, , Leipzig, Germany|
|Date Published||16 - 18 July|
Automatically generated, human-quality text summarization systems are difficult both to develop and to evaluate, partly because articles differ along several dimensions: length, writing style and lexical usage. In this paper we propose a framework that, by utilizing RSS feeds, is able to personalize on the needs of the users and on the needs of their device, in order to present to the end-user only a fraction of the news articles covering just the useful information that derives from them. The created summaries utilize a weighted combination of statistical and linguistic features which leads to sentence scoring and selection. The procedure is assisted by categorization results as well as personalization algorithms that enhance the summarization module. The mechanism is evaluated using classic precision-recall metrics together with statistical results from real users. Within this framework we have created the PeRSSonal system that is able to create personalized, pre-categorized, dynamically generated RSS feeds focalized on the end user?s small screen device.