### Abstract

This paper describes CMIC's submissions to the TREC'09 relevance feedback track. In the phase 1 runs we submitted, we experimented with two different techniques to produce 5 documents to be judged by the user in the initial feedback step, namely using knowledge bases and clustering. Both techniques attempt to topically diversify these 5 documents as much as possible in an effort to maximize the probability that they contain at least 1 relevant document. The basic premise is that if a query has n diverse interpretations, then diversifying results and picking the top 5 most likely interpretations would maximize the probability that a user would be interested in at least one interpretation. In phase 2 runs, which involved the use of the feedback attained from phase 1 judgments, we attempted to use positive and negative judgments in weighing the terms to be used for subsequent feedback.

Original language | English |
---|---|

Title of host publication | NIST Special Publication |

Publication status | Published - 2009 |

Externally published | Yes |

Event | 18th Text REtrieval Conference, TREC 2009 - Gaithersburg, MD, United States Duration: 17 Nov 2009 → 20 Nov 2009 |

### Other

Other | 18th Text REtrieval Conference, TREC 2009 |
---|---|

Country | United States |

City | Gaithersburg, MD |

Period | 17/11/09 → 20/11/09 |

### Fingerprint

### ASJC Scopus subject areas

- Engineering(all)

### Cite this

*NIST Special Publication*

**CMIC@TREC-2009 : Relevance feedback track.** / Darwish, Kareem; El-Deeb, Ahmed.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

*NIST Special Publication.*18th Text REtrieval Conference, TREC 2009, Gaithersburg, MD, United States, 17/11/09.

}

TY - GEN

T1 - CMIC@TREC-2009

T2 - Relevance feedback track

AU - Darwish, Kareem

AU - El-Deeb, Ahmed

PY - 2009

Y1 - 2009

N2 - This paper describes CMIC's submissions to the TREC'09 relevance feedback track. In the phase 1 runs we submitted, we experimented with two different techniques to produce 5 documents to be judged by the user in the initial feedback step, namely using knowledge bases and clustering. Both techniques attempt to topically diversify these 5 documents as much as possible in an effort to maximize the probability that they contain at least 1 relevant document. The basic premise is that if a query has n diverse interpretations, then diversifying results and picking the top 5 most likely interpretations would maximize the probability that a user would be interested in at least one interpretation. In phase 2 runs, which involved the use of the feedback attained from phase 1 judgments, we attempted to use positive and negative judgments in weighing the terms to be used for subsequent feedback.

AB - This paper describes CMIC's submissions to the TREC'09 relevance feedback track. In the phase 1 runs we submitted, we experimented with two different techniques to produce 5 documents to be judged by the user in the initial feedback step, namely using knowledge bases and clustering. Both techniques attempt to topically diversify these 5 documents as much as possible in an effort to maximize the probability that they contain at least 1 relevant document. The basic premise is that if a query has n diverse interpretations, then diversifying results and picking the top 5 most likely interpretations would maximize the probability that a user would be interested in at least one interpretation. In phase 2 runs, which involved the use of the feedback attained from phase 1 judgments, we attempted to use positive and negative judgments in weighing the terms to be used for subsequent feedback.

UR - http://www.scopus.com/inward/record.url?scp=84873452007&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84873452007&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84873452007

BT - NIST Special Publication

ER -