Browsing and Matching − Scoping Status: Final Report Overview of the Current Situation: Matching the Multilingual Approach in Today's Search Engines Universal Character Set

1 Executive summary Today's society is on its way from a traditionally production-based economy to a knowledge-based economy. The European Commission's action plan on Europe's way to the information society 1 outlines some of the major developments in this field and recommends steps to be undertaken to prepare Europe for this challenge. Obviously, the Information Society is not only about information, not even only about access to information, it is also about locating relevant information. In many ways, information retrieval is the Web revolution's neglected child. Even the otherwise excellent Information Society Glossary 2 does not refer to this crucial topic. Of course, search engines, portal sites, and indexing services do exist. However, in contrast to many of the other topics in this field, the question of locating information involves not only international standards, but also specifically European, national, regional, social, and even personal factors. Many of these issues are related to Europe's multilingual and multicultural heritage which European institutions, including standards bodies such as CEN/TC304 »European localization requirements«, must strive to protect. The issues encompass points such as: − Existence of relevant information in many languages; − The use of different scripts (e. g. Latin, Greek, and Cyrillic scripts); − The use of letters which are particular to a given language or a number of languages; − Expectations how such letters or scripts are handled in more restricted character sets such as ASCII (fallback, transliteration, input methods); − Familiarity with certain cataloguing schemes / database categories specific to a country / a group of countries. The task soon becomes more ambitious. Human readers 3 will naturally recognize that sing, sang, sung 4 are just three tenses of the very same verb, just as oeil and yeux differ only with respect to number. They will also not mix the German word Boot with its English homograph of completely different meaning, 5 whereas they understand at once that Pericles, Perikles and Περικλη Ä ς are really one and the same person 6 and that browsing and scanning can be synonyms 7 in some contexts but not in others. 8 For English, with its fairly limited number of irregular verbs and its otherwise rather regular construction of derived forms, some of these problems can still be dealt with relatively easily in comparison with most other European languages where word formation is more complex. While no speedy solution is to be expected, these …