Our paper from the IMAIL-project with the title Comparing Manual Text Patterns and Machine Learning for Classification of E-Mails for Automatic Answering by a Government Agency by H. Dalianis, J. Sjöbergh and E. Sneiders, were presented at CICLing 2011, 12th International Conference on Intelligent Text Processing and Computational Linguistics, February 20-26, Tokyo, Japan, by Jonas Sjöbergh who participates in the IMAIL project from KTH. The proceedings were printed in Springer Verlag.
The paper shows that a manual approach to create rules for answering e-mails (in Swedish) is more accurate than using a machine learning approach, though a machine learning approach gives higher recall.
CICLing had around 100 participants below a couple of photos.
