Abstract | ||
---|---|---|
Using language technology for text analysis and light-weight ontologies as a content-mediating level, we acquire indexing patterns from vast amounts of indexing data for English-language medical documents. This is achieved by statistically relating interlingual representations of these documents (based on text token bigrams) to their associated index terms. From these 'English' indexing patterns, we then induce the associated index terms for German and Portuguese documents when their interlingual representations match those of English documents. Thus, we learn from past English indexing experience and transfer it in an unsupervised way to non-English texts, without ever having seen concrete indexing data for languages other than English. |
Year | Venue | Keywords |
---|---|---|
2004 | AAAI | associated index term,text analysis,concrete indexing data,past english indexing experience,english document,interlingual representation,text token bigrams,indexing pattern,english-language medical document,indexing data,indexation,english language,language technology,indexing terms |
Field | DocType | ISBN |
Ontology (information science),Medical documents,Computer science,Portuguese,Search engine indexing,Natural language processing,Bigram,Artificial intelligence,Security token,Language technology,German | Conference | 0-262-51183-5 |
Citations | PageRank | References |
1 | 0.36 | 6 |
Authors | ||
3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Udo Hahn | 1 | 88 | 11.14 |
Kornél Markó | 2 | 38 | 3.58 |
Stefan Schulz | 3 | 29 | 5.15 |