Deidentification of clinical records is a crucial step before these records can be distributed to non-hospital researchers. Most approaches to deidentification rely heavily on dictionaries and heuristic rules; these approaches fail to remove most personal health information (PHI) that cannot be found in dictionaries. They also can fail to remove PHI that is ambiguous between PHI and non-PHI. Na...

This paper proposes a named entity recognition (NER) method for speech recognition results that uses confidence on automatic speech recognition (ASR) as a feature. The ASR confidence feature indicates whether each word has been correctly recognized. The NER model is trained using ASR results with named entity (NE) labels as well as the corresponding transcriptions with NE labels. In experiments...

This paper presents a method for the named entity recognition (NER) of speech data that uses automatic speech recognition (ASR) confidence as a feature that indicates whether each word is correctly recognized. An NER model is trained using ASR results with named entity (NE) labels to include an ASR confidence feature as well as corresponding transcriptions with NE labels. Experiments using supp...

This paper presents an approach that uses structural information for Japanese named entity recognition (NER). Our NER system is based on Support Vector Machine (SVM), and utilizes four types of structural information: cache features, coreference relations, syntactic features and caseframe features, which are obtained from structural analyses. We evaluated our approach on CRL NE data and obtaine...

Named entity recognition is a fundamental task in biomedical data mining. Multiple -class annotation is more challenging than single class annotation. In this paper, we took a single word classification approach to dealing with the multiple -class annotation problem using Support Vector Machines (SVMs). Word attributes, results of existing gene/protein name taggers, context, and other informati...

This paper describes a hybrid model and the corresponding algorithm combining support vector machines (SVMs) with statistical methods to improve the performance of SVMs for the task of Chinese Named Entity Recognition (NER). In this algorithm, a threshold of the distance from the test sample to the hyperplane of SVMs in feature space is used to separate SVMs region and statistical method region...