TheAsiaInformationRetrievalSymposium(AIRS)wasestablishedbytheAsian information retrieval community after the successful series of Information - trieval with Asian Languages (IRAL) workshops held in six di?erent locations in Asia, starting from 1996. While the IRAL workshops had their focus on inf- mation retrieval problems involving Asian languages, AIRS covers a wider scope of applications, systems, technologies and theory aspects of information retrieval in text, audio, image, video and multimedia data. This extension of the scope re?ects and fosters increasing research activities in information retrieval in this region and the growing need for collaborations across subdisciplines. We are very pleased to report that we saw a sharp increase in the number of submissions and their quality, compared to the IRAL workshops. We received 106papersfromninecountriesinAsiaandNorthAmerica,fromwhich28papers (26%) were presented in oral sessions and 38 papers in poster sessions (36%). It was a great challenge for the Program Committee to select the best among the excellent papers. The low acceptance rates witness the success of this year's conference. After a long discussion between the AIRS 2004 Steering Committee and Springer, the publisher agreed to publish our proceedings in the Lecture Notes in Computer Science (LNCS) series, which is SCI-indexed. We feel that this strongly attests to the excellent quality of the papers.

Inhalt
Information Organization.- Automatic Word Clustering for Text Categorization Using Global Information.- Text Classification Using Web Corpora and EM Algorithms.- Applying CLIR Techniques to Event Tracking.- Document Clustering Using Linear Partitioning Hyperplanes and Reallocation.- Automatic Summarization.- Summary Generation Centered on Important Words.- Sentence Compression Learned by News Headline for Displaying in Small Device.- Automatic Text Summarization Using Two-Step Sentence Extraction.- Sentence Extraction Using Time Features in Multi-document Summarization.- Alignment/Paraphrasing in IR.- Extracting Paraphrases of Japanese Action Word of Sentence Ending Part from Web and Mobile News Articles.- Improving Transliteration with Precise Alignment of Phoneme Chunks and Using Contextual Features.- Combining Sentence Length with Location Information to Align Monolingual Parallel Texts.- Web Search.- Effective Topic Distillation with Key Resource Pre-selection.- Efficient PageRank with Same Out-Link Groups.- Literal-Matching-Biased Link Analysis.- Linguistic Issues in IR.- Multilingual Relevant Sentence Detection Using Reference Corpus.- A Bootstrapping Approach for Geographic Named Entity Annotation.- Using Verb Dependency Matching in a Reading Comprehension System.- Document/Query Models.- Sense Matrix Model and Discrete Cosine Transform.- Query Model Estimations for Relevance Feedback in Language Modeling Approach.- A Measure Based on Optimal Matching in Graph Theory for Document Similarity.- Estimation of Query Model from Parsimonious Translation Model.- Enabling Technology.- Ranking the NTCIR Systems Based on Multigrade Relevance.- X-IOTA: An Open XML Framework for IR Experimentation.- Recognition-Based Digitalization of Korean Historical Archives.- On Bit-Parallel Processing of Multi-byte Text.- Mobile Applications.- Retrieving Regional Information from Web by Contents Localness and User Location.- Towards Understanding the Functions of Web Element.- Clustering-Based Navigation of Image Search Results on Mobile Devices.