Combining Multiple Strategies for Effective Monolingual and Cross-Language Retrieval

Abstract

Abstract. This paper describes and evaluates different retrieval strategies that are useful for search operations on document collections written in various European languages, namely French, Italian, Spanish and German. We also suggest and evaluate different query translation schemes based on freely available translation resources. In order to cross language barriers, we propose a combined query translation approach that has resulted in interesting retrieval effectiveness. Finally, we suggest a collection merging strategy based on logistic regression that tends to perform better than other merging approaches.

Ad Hoc Retrieval with the Persian Language

“…Each run is based on a data fusion scheme combining several single runs using different IR models (DFR, Okapi, and language model (LM)), indexing strategies (word with and without stemming or 5-gram), query expansion strategies (Rocchio, idf -based or none) and query formulation (T, TD, and TDN). The fusion was performed for all four runs using our Z-score operator [18]. In all cases we can see that combining different models, indexing and search strategies using Z-score approach improves clearly the retrieval effectiveness.…”

Selection and Merging Strategies for Multilingual Information Retrieval

“…On the other hand, differences between the various translation tools are usually not statistically significant, except for a few such as "Babylon 2" and "Babylon 3" for both French and Portuguese, or "InterTran" for French and Russian. It is known that although a given translation tool may produce acceptable translations for a given set of requests, it may perform poorly for other queries [1], [2]. To date we have not been able to detect with much precision when a given translation will produce satisfactory retrieval performance and when it will fail.…”

“…Based on our bilingual and multilingual experiments of the last years [1], [2], we conducted additional experiments involving various bilingual and multilingual test-collections. Based on a request written in English, we retrieved documents written in English, French, Finnish and Russian.…”

Monolingual, Bilingual, and GIRT Information Retrieval at CLEF-2005

“…Since 2001, our research group has been investigating effective information retrieval (IR) techniques capable of handling a variety of natural languages [1], [2], in order to improve both monolingual and bilingual searches. Along this same stream, and based on our assumption that nouns and adjectives reveal the most about semantic content of documents (or requests), we designed a set of stopword lists and light stemming procedures for certain European languages.…”