Title | ||
---|---|---|
Unsupervised Language Model Adaptation using Utterance-based Web Search for Clinical Speech Recognition. |
Abstract | ||
---|---|---|
In this working notes paper we present our methodology in clinical speech recognition for the Task 1.a.1 of the CLEF eHealth Evaluation Lab 2015. The goal of this task is to minimize the word- detection errors. Our approach is based on the assumption that each spoken clinical document has its own context. Hence, the recognition system is adapted for each document separately. The proposed method performs two-pass decoding whereas the rst transcript is processed to queries which are used for retrieving web resources as adaptation data to build a document-specic dictionary and language model. The second pass decodes the same document using the adapted dictionary and lan- guage model. The experimental results show a reduction of the insertion errors in comparison to the baseline system, but no improvement of the overall incorrectness percentage across all spoken documents. |
Year | Venue | Field |
---|---|---|
2015 | CLEF (Working Notes) | Web resource,Computer science,Utterance,Speech recognition,eHealth,Natural language processing,Artificial intelligence,Baseline system,Decoding methods,Language model,Clef,Decodes |
DocType | Citations | PageRank |
Conference | 1 | 0.35 |
References | Authors | |
8 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Robert Herms | 1 | 5 | 3.90 |
Daniel Richter | 2 | 60 | 9.52 |
Maximilian Eibl | 3 | 119 | 37.66 |
Marc Ritter | 4 | 21 | 15.52 |