Thursday, September 3, 2009
Good news from ISDA 2009 and VLDL 2009
Returning from vacation I have been greeted not by one, but two good news.
First, a paper from Stefano Baccianella, Fabrizio Sebastiani, and me has been accepted at the International Conference on Intelligent Systems Design and Applications (ISDA), which will be held in Pisa, Italy (yes, where I live).
The paper presents a study on Evaluation Measures for Ordinal Text Classification, proposing a measure of Macro-Mean Average Error, which we think is better suited (with respect to the usually adopted mean average error measure) to evaluate ordinal regression experiments in case on highly unbalanced data distribution.
The presence of unbalanced data distribution is a relatively common case in ordinal text classification, e.g., star-like scores assigned to products in products reviewers are typically highly skewed toward all-star ratings.
Second, a paper from Paolo Bolettieri, Fabrizio Falchi, Claudio Lucchese, Raffaele Perego, Fausto Rabitti and me has been accepted at the Second Workshop on Very Large Digital Libraries (VLDL), which will be held in Corfu, Greece.
The paper summarizes the relevant aspects of experience of building the CoPhIR collection.
Add comment
Fill out the form below to add your own comments