Cross-language information retrieval method based on conceptual graph

Abstract

The invention discloses a cross-language information retrieval method based on a conceptual graph. The concept map-based cross-language information retrieval method is a cross-language similarity evaluation framework CG-CLIR method, which comprises the following steps: S1, based on the concept edge representation of Skip-Gram; S2, high-order semantic embedding and similarity calculation of the fusion side information; the step S1 includes text preprocessing, concept graph edge semantic embedding. The step S2 includes LSTM based graph level semantic embedding, graph embedding based similarity calculation. According to the method, in text cross-language information retrieval, language obstacles are crossed, and the semantic retrieval effect is achieved under the condition that translation isnot conducted.

Description

technical field [0001] The invention relates to cross-language information retrieval, in particular to a cross-language information retrieval method based on concept maps. Background technique [0002] At present, vector representation and processing of text has become the mainstream of text analysis tasks. The most common one is to vectorize the representation of words, such as the one-hot model and word embedding model, which represent n words in a sentence as n A d-dimensional vector, so that the sentence is an n*d-dimensional matrix, which is convenient for processing. Another method is to map sentences or documents into a vector, and construct paragraphs and texts into a matrix of vectors. When dealing with this method, more consideration is given to longer sequence information issues, and it can better represent global information. However, since the sentence is variable in length, and as the basic representation unit of semantics, words can have many different combi...

AI Technical Summary

Problems solved by technology

[0004] Traditional retrieval methods often use shallow text feature information to evaluate similarity. Even if semantics is used for similarity comparison, it is often only reflected at the word level, and the use of global phonetic features is not sufficient.

In the cross-language similarity retrieval task, due to the differences in the grammar of different languages, the translation of words also has the problem of disambiguation, and generally cannot be retrieved by simple text features.

Images