Sensitive documents are those that cannot be made public, e.g., for personal or organizational privacy reasons. For instance, documents requested through Freedom of Information mechanisms must be manually reviewed for the presence of sensitive information before their actual release. Hence, tools that can assist human reviewers in spotting sensitive information are of great value to government organizations subject to Freedom of Information laws. We look at sensitivity identification in terms of semi-automated text classification (SATC), the task of ranking automatically classified documents so as to optimize the cost-effectiveness of human post-checking work. We use a recently proposed utility-theoretic approach to SATC that explicitly optimizes the chosen effectiveness function when ranking the documents by sensitivity; this is especially useful in our case, since sensitivity identification is a recall-oriented task, thus requiring the use of a recall-oriented evaluation measure such as F2. We show the validity of this approach by running experiments on a multi-label multi-class dataset of government documents manually annotated according to different types of sensitivity.
Semi-automated text classification for sensitivity identification
Esuli A;Sebastiani F
2015
Abstract
Sensitive documents are those that cannot be made public, e.g., for personal or organizational privacy reasons. For instance, documents requested through Freedom of Information mechanisms must be manually reviewed for the presence of sensitive information before their actual release. Hence, tools that can assist human reviewers in spotting sensitive information are of great value to government organizations subject to Freedom of Information laws. We look at sensitivity identification in terms of semi-automated text classification (SATC), the task of ranking automatically classified documents so as to optimize the cost-effectiveness of human post-checking work. We use a recently proposed utility-theoretic approach to SATC that explicitly optimizes the chosen effectiveness function when ranking the documents by sensitivity; this is especially useful in our case, since sensitivity identification is a recall-oriented task, thus requiring the use of a recall-oriented evaluation measure such as F2. We show the validity of this approach by running experiments on a multi-label multi-class dataset of government documents manually annotated according to different types of sensitivity.File | Dimensione | Formato | |
---|---|---|---|
prod_344510-doc_107919.pdf
solo utenti autorizzati
Descrizione: Semi-automated text classification for sensitivity identification
Tipologia:
Versione Editoriale (PDF)
Dimensione
340.28 kB
Formato
Adobe PDF
|
340.28 kB | Adobe PDF | Visualizza/Apri Richiedi una copia |
prod_344510-doc_156927.pdf
accesso aperto
Descrizione: Semi-automated text classification for sensitivity identification
Tipologia:
Versione Editoriale (PDF)
Dimensione
102.17 kB
Formato
Adobe PDF
|
102.17 kB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.