| other,17-1-N04-4028,bq | </term> , such as the <term> Web </term> or <term> | newswire documents | </term> . Despite the successes of these | #6771 Information extraction techniques automatically create structured databases from unstructured data sources, such as the Web ornewswire documents. |
| tech,19-4-N04-4028,bq | conditional random field ( CRF ) </term> , a <term> | probabilistic model | </term> which has performed well on <term> | #6830 The information extraction system we evaluate is based on a linear-chain conditional random field (CRF), aprobabilistic model which has performed well on information extraction tasks because of its ability to capture arbitrary, overlapping features of the input in a Markov model. |
| other,26-4-N04-4028,bq | </term> which has performed well on <term> | information extraction tasks | </term> because of its ability to capture | #6837 The information extraction system we evaluate is based on a linear-chain conditional random field (CRF), a probabilistic model which has performed well oninformation extraction tasks because of its ability to capture arbitrary, overlapping features of the input in a Markov model. |
| other,10-5-N04-4028,bq | the <term> confidence </term> of both <term> | extracted fields | </term> and entire <term> multi-field records | #6868 We implement several techniques to estimate the confidence of bothextracted fields and entire multi-field records, obtaining an average precision of 98% for retrieving correct fields and 87% for multi-field records. |
| other,7-5-N04-4028,bq | several techniques to estimate the <term> | confidence | </term> of both <term> extracted fields </term> | #6865 We implement several techniques to estimate theconfidence of both extracted fields and entire multi-field records, obtaining an average precision of 98% for retrieving correct fields and 87% for multi-field records. |
| other,14-5-N04-4028,bq | <term> extracted fields </term> and entire <term> | multi-field records | </term> , obtaining an <term> average precision | #6872 We implement several techniques to estimate the confidence of both extracted fields and entiremulti-field records, obtaining an average precision of 98% for retrieving correct fields and 87% for multi-field records. |
| measure(ment),19-5-N04-4028,bq | multi-field records </term> , obtaining an <term> | average precision | </term> of 98 % for retrieving correct <term> | #6877 We implement several techniques to estimate the confidence of both extracted fields and entire multi-field records, obtaining anaverage precision of 98% for retrieving correct fields and 87% for multi-field records. |
| other,5-1-N04-4028,bq | techniques </term> automatically create <term> | structured databases | </term> from <term> unstructured data sources | #6759 Information extraction techniques automatically createstructured databases from unstructured data sources, such as the Web or newswire documents. |
| other,12-3-N04-4028,bq | desirable to accurately estimate the <term> | confidence | </term> the system has in the correctness | #6799 For many reasons, it is highly desirable to accurately estimate theconfidence the system has in the correctness of each extracted field. |
| tech,10-4-N04-4028,bq | system </term> we evaluate is based on a <term> | linear-chain conditional random field ( CRF ) | </term> , a <term> probabilistic model </term> | #6821 The information extraction system we evaluate is based on alinear-chain conditional random field ( CRF ), a probabilistic model which has performed well on information extraction tasks because of its ability to capture arbitrary, overlapping features of the input in a Markov model. |
| other,15-1-N04-4028,bq | unstructured data sources </term> , such as the <term> | Web | </term> or <term> newswire documents </term> | #6769 Information extraction techniques automatically create structured databases from unstructured data sources, such as theWeb or newswire documents. |
| other,21-3-N04-4028,bq | system has in the correctness of each <term> | extracted field | </term> . The <term> information extraction | #6808 For many reasons, it is highly desirable to accurately estimate the confidence the system has in the correctness of eachextracted field. |
| other,32-5-N04-4028,bq | correct <term> fields </term> and 87 % for <term> | multi-field records | </term> . We present a novel approach for | #6890 We implement several techniques to estimate the confidence of both extracted fields and entire multi-field records, obtaining an average precision of 98% for retrieving correct fields and 87% formulti-field records. |
| other,8-1-N04-4028,bq | <term> structured databases </term> from <term> | unstructured data sources | </term> , such as the <term> Web </term> or <term> | #6762 Information extraction techniques automatically create structured databases fromunstructured data sources, such as the Web or newswire documents. |
| tech,0-1-N04-4028,bq | baseline </term> on all three aspects . <term> | Information extraction techniques | </term> automatically create <term> structured | #6754 Results indicate that the system yields higher performance than a baseline on all three aspects.Information extraction techniques automatically create structured databases from unstructured data sources, such as the Web or newswire documents. |
| other,38-4-N04-4028,bq | to capture arbitrary , overlapping <term> | features | </term> of the input in a <term> Markov model | #6849 The information extraction system we evaluate is based on a linear-chain conditional random field (CRF), a probabilistic model which has performed well on information extraction tasks because of its ability to capture arbitrary, overlappingfeatures of the input in a Markov model. |
| measure(ment),7-2-N04-4028,bq | Despite the successes of these systems , <term> | accuracy | </term> will always be imperfect . For many | #6781 Despite the successes of these systems,accuracy will always be imperfect. |
| other,27-5-N04-4028,bq | </term> of 98 % for retrieving correct <term> | fields | </term> and 87 % for <term> multi-field records | #6885 We implement several techniques to estimate the confidence of both extracted fields and entire multi-field records, obtaining an average precision of 98% for retrieving correctfields and 87% for multi-field records. |
| model,44-4-N04-4028,bq | <term> features </term> of the input in a <term> | Markov model | </term> . We implement several techniques | #6855 The information extraction system we evaluate is based on a linear-chain conditional random field (CRF), a probabilistic model which has performed well on information extraction tasks because of its ability to capture arbitrary, overlapping features of the input in aMarkov model. |
| tech,1-4-N04-4028,bq | each <term> extracted field </term> . The <term> | information extraction system | </term> we evaluate is based on a <term> linear-chain | #6812 Theinformation extraction system we evaluate is based on a linear-chain conditional random field (CRF), a probabilistic model which has performed well on information extraction tasks because of its ability to capture arbitrary, overlapping features of the input in a Markov model. |
