COLING 2008 =========== COLING 2008, the 22nd International Conference on Computational Linguistics was held in Manchester, UK in August. It was the first COLING conference in the UK, a country with a rich history and lively research scene in Computational Linguistics. The great response to the call for papers may have been caused by this location or it may just have been a consequence of the rapid growth of our discipline. In any case, the 600 submissions of high average quality we received made it relatively easy for the program committee to put together an excellent program ( see http://www.coling2008.org.uk/). The Programme: COLING'08 attracted 600 submissions of papers, posters and demos. Of these 48 were not reviewed because they were either withdrawn or rejected for failing to meet the the submission requirements. The geographical profile of the remaining submissions was: 91 United States 90 Japan 60 China 53 United Kingdom 41 Germany 32 France 18 Spain 16 India 13 Taiwan Province of China 12 Republic of Korea 11 Canada, Australia 10 Italy 9 Netherlands, Israel, Ireland 8 Hong Kong Special Administrative Region of China 7 Singapore 6 Switzerland, Belgium 5 Thailand, Sweden 3 Romania, Portugal 2 Macau, Greece, Finland, Brazil 1 Turkey, South Africa, Slovenia, Russian Federation, Poland, Mexico, Malaysia, Luxembourg, Lithuania, Jordan, Iran (Islamic Republic of), Denmark, Czech Republic, Cuba, Bulgaria, Bangladesh After a thorough reviewing process including a period of interactive deliberation, the program committee selected 146 full papers and 36 poster presentations. The central criterion for the selection was scientific quality rather than geographic balance or the desirable spread across subareas. We tried to apply a multidimensional concept of quality that does not exclusively favour technically sound engineering papers but also yields some space for challenging scientific insights and first reports on novel approaches. Looking at the distribution of the papers among subfields of CL, we made a few observations. One concerns the central theme of machine learning. Although the term machine learning only appears in the name of one single session, machine learning actually transcends nearly all represented subfields of our discipline. After decades of hibernation, the area of machine translation has again become a central field of research. Almost all of the MT related submissions are on statistical translation but a growing number of papers describes clever combinations of mehods from different paradigms. Compared with MT, the area of natural language generation is much less represented, which may partially be due to this year's International Language Generation Conference in Ohio. The area of information extraction still keeps growing. With subareas such as opinion mining, sentiment detection and event extraction it has become rather diversified. A special observation concerns specialised types of phrase disambiguation or classification that cannot easily be subsumed under IR or IE since the described methods could also be utilised for summarisation, paraphrasing or other application types. In general it has become harder to assign method papers to just one traditional technology area. This is nicely reflected in the authors' choice of multiple keywords from different areas. We only received few submissions on speech technologies, in our opinion even less than in earlier COLING conferences. Although this development might simply be attributed to the inevitable and ever progressing differentiation of the human language technologies, it may also be the case that the meeting market in this area is well covered by the well known speech conferences. This year's ACL conference also had just a single speech processing paper. We hope that our colleagues will forgive us for having been rather strict with double submissions. In several cases accepted submissions were finally turned down because a paper with largely overlapping content had appeared or was scheduled to appear elsewhere (We also had to turn down a number of papers that were not annonymised). We believe that our field has to find a proper way of dealing with the increasing number of professional conferences without sacrificing the basic principles of scientific publishing. We were priviliged to have keynote addresses by two leading scientists in related areas: Elizabeth Shriberg (SRI and Berkeley) on Challenges to Using Prosody in Automatic Language Processing and John Shawe-Taylor (UCL) on Machine Learning for multimodal analysis. After deciding to enrich this Conference with a Best COLING Paper Award, we received an offer from the renowned scientific publishing house Springer to support such and award. We are grateful to Springer for this generous donation and thank especially Olga Chiarcos for her efforts in this case. The Best COLING Paper Award went to Bill MacCartney and Christopher Manning for their paper on "Modeling Semantic Containment and Exclusion in Natural Language Inference". We encourage you all to read it, and the other papers. The COLING 2008 proceedings are already included in the ACL Anthology! Together with Olga Chiarcos we also thought about means for making COLING even more attractive and visible. Olga proposed a special book publication of selected ground-breaking COLING papers in their extended versions. This is an excellent idea which is being implemented already for this COLING conference. The geographical profile of the accepted submissions was: 62 Japan 48 United States 43 China 31 United Kingdom 28 Germany 20 France 13 Spain 12 Republic of Korea 11 Taiwan Province of China, India 8 Italy 6 Israel, Canada, Australia 5 Switzerland, Hong Kong Special Administrative Region of China 4 Thailand, Ireland, Belgium 3 Sweden, Singapore, Portugal, Netherlands 2 Romania, Macau, Greece, Brazil 1 Turkey, South Africa, Slovenia, Russian Federation, Mexico, Malaysia, Luxembourg, Lithuania, Jordan, Iran (Islamic Republic of), Cuba, Bulgaria, Bangladesh Attendees: The conference attracted 421 attendees from 37 countries. The breakdown for this is: USA 85 UK 64 Japan 48 China 38 France 32 Germany 26 Netherlands 17 Spain 16 Ireland 15 Korea 12 Canada 10 Sweden 9 Russia 8 Italy 8 Switzerland 7 India 7 Belgium 6 Australia 6 Taiwan 5 Finland 5 Czech Rep 5 Singapore 4 Romania 4 Israel 3 Thailand 2 Portugal 2 Norway 2 Denmark 2 Syria 1 Poland 1 Mexico 1 Malta 1 Hungary 1 Egypt 1 Brazil 1 Austria 1 Vietnam 1 We want to thank the area chairs who steered the reviewing process to a fruitful end, and to acknowledge the successful work of our numerous colleagues who acted as reviewers. Our special gratitude goes to Roger Evans for his hard and uncompromising work on the proceedings and Christian Spurk for the excellent technical support he gave to the Programme Committee. We would also like to thank the local organizer Harold Somers for providing such a good venue, outing and banquet, for his hard work in achieving this, and for his constructive cooperation with the Programme Committee. COLING 2010 will be in Beijing. We look forward to seeing you there! Donia Scott and Hans Uszkoreit Program Co-Chairs of COLING 2008