| A00-2004 | compare three versions of the | TextTiling | algorithm ( Hearst , 1994 ) . |
| D11-1026 | document . A classical approach is | TextTiling | ( Hearst , 1997 ) . It consists |
| C02-1006 | segmentation algorithm is better than | TextTiling | . But the improvement is still |
| J05-2005 | Unlike approaches such as the | TextTiling | algorithm ( Hearst 1997 ) , ours |
| E06-1035 | LCSeg algorithm , a variant of | TextTiling | ( Hearst , 1997 ) . LCSeg hypothesizes |
| J02-4003 | of Hearst 's -LSB- 1997 -RSB- | TextTiling | algorithm . ) The three components |
| C02-1006 | the documents . After through | TextTiling | algorithm , a file will be broken |
| C02-1006 | tiling approach ( Hearst , 1993 ) . | TextTiling | subdivides text into multi-paragraph |
| J02-4008 | lexical chains , but not as well as | TextTiling | . Hearst also appears as a computational |
| A00-2004 | b ) ) ( 9 ) 4.3 Experiment 2 - | TextTiling | We compare three versions of |
| D10-1038 | boundary . This method is similar to | TextTiling | ( Hearst , 1997 ) except that |
| J02-4008 | chains and Hearst 's ( 1997 ) | TextTiling | on a corpus of 300 texts . ( |
| D13-1130 | strategy of local methods such as | TextTiling | ( Hearst , 1997 ) . The cosine |
| C00-1072 | segmentation algorithms such as | TextTiling | ( Hearst , t997 ) to segment |
| D08-1001 | Hearst ( 1997 ) , who proposed the | Texttiling | algorithm . The best results |
| C02-1033 | results than a system such as | TextTiling | that is based on word recurrence |
| E06-1035 | , term repetitions ) , whereas | TextTiling | computes similarity using word |
| E06-1035 | principal difference between LCSeg and | TextTiling | is that LCSeg measures similarity |
| J09-1006 | based on Hearst 's ( 1994 , 1997 ) | TextTiling | procedure . TextTiling is a quantitative |
| D08-1079 | subtopic passages by using the | TextTiling | algorithm ( Hearst , 1997 ) , |
