| Project/Area Number |
26330337
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Research Field |
Life / Health / Medical informatics
|
|---|
| Research Institution | Kansai Medical University |
|---|
Principal Investigator |
WATANABE Jun 関西医科大学, 医学部, 准教授 (40148557)
|
|---|
| Project Period (FY) |
2014-04-01 – 2017-03-31
|
| Project Status |
Completed (Fiscal Year 2016)
|
| Budget Amount *help |
¥1,690,000 (Direct Cost: ¥1,300,000、Indirect Cost: ¥390,000)
Fiscal Year 2016: ¥520,000 (Direct Cost: ¥400,000、Indirect Cost: ¥120,000)
Fiscal Year 2015: ¥650,000 (Direct Cost: ¥500,000、Indirect Cost: ¥150,000)
Fiscal Year 2014: ¥520,000 (Direct Cost: ¥400,000、Indirect Cost: ¥120,000)
|
|---|
| Keywords | 日本語自然文 / 電子カルテ / 構文解析 / テキストマイニング / 正規化変換 / パラレルコーパス / 診療録 / アルゴリズム / 自然言語処理 / 日本語正規変換 |
|---|
| Outline of Final Research Achievements |
An algorithm to translate natural Japanese sentences into computer processable and human understandable "normalized" sentences was designed and verified the translation ability of the algorithm using the descriptions in the assessment and plan sections of electric medical records. The descriptions were anonymized and subjected to the syntactic analysis followed by the rule-based translation with text mining techniques to convert natural sentences into the normalized sentences with correct syntax and no ambiguity. A parallel corpus consists of natural- and normalized-sentence entries thus formed was then used to convert the remaining natural sentences. About 95% of the natural sentences could be converted into the normalized sentences along the algorithm. When the algorithm is modified slightly, the modified algorithm is applicable to the supervised machine learning, suggesting that the accuracy of the machine learning improve by the introduction of the modified algorithm.
|
