This paper presents a method for opinion lexicon expansion and target extraction through double propagation. The method uses syntactic relations between opinion words and targets, identified via dependency parsing, to iteratively expand the initial opinion lexicon and extract targets. The approach is semi-supervised, requiring only an initial seed opinion lexicon. It propagates information between opinion words and targets, enabling the extraction of new opinion words and targets. The method is evaluated against state-of-the-art approaches on a product review dataset, showing significant improvements in performance. The key contributions of the paper include: (1) a novel propagation-based method for opinion lexicon expansion and target extraction, (2) the use of syntactic relations to identify and propagate information between opinion words and targets, and (3) the development of methods for assigning sentiment polarities to newly extracted opinion words and pruning noisy targets. The method is based on the observation that opinion words are used to modify targets, and that opinion words and targets themselves have relations in opinionated expressions. These relations can be identified via a dependency parser and then exploited to perform the extraction tasks. The propagation process iteratively extracts opinion words and targets using known and extracted opinion words and targets through the identification of syntactic relations. The identification of the relations is the key to the extractions. The method is evaluated on a product review dataset, and the results show that the approach outperforms existing methods in both opinion lexicon expansion and target extraction. The method is also effective in assigning sentiment polarities to newly extracted opinion words and pruning noisy targets. The results demonstrate that the proposed method is effective in extracting a large number of new opinion words and targets with high precision and recall. The method is also effective in handling different numbers of seeds, with the best performance achieved when using a moderate number of seeds. The method is also effective in handling different types of opinionated expressions, including those with negations and contraries. The method is also effective in handling different types of product reviews, including those with complex syntactic structures. The method is also effective in handling different types of opinionated expressions, including those with negations and contraries. The method is also effective in handling different types of product reviews, including those with complex syntactic structures.This paper presents a method for opinion lexicon expansion and target extraction through double propagation. The method uses syntactic relations between opinion words and targets, identified via dependency parsing, to iteratively expand the initial opinion lexicon and extract targets. The approach is semi-supervised, requiring only an initial seed opinion lexicon. It propagates information between opinion words and targets, enabling the extraction of new opinion words and targets. The method is evaluated against state-of-the-art approaches on a product review dataset, showing significant improvements in performance. The key contributions of the paper include: (1) a novel propagation-based method for opinion lexicon expansion and target extraction, (2) the use of syntactic relations to identify and propagate information between opinion words and targets, and (3) the development of methods for assigning sentiment polarities to newly extracted opinion words and pruning noisy targets. The method is based on the observation that opinion words are used to modify targets, and that opinion words and targets themselves have relations in opinionated expressions. These relations can be identified via a dependency parser and then exploited to perform the extraction tasks. The propagation process iteratively extracts opinion words and targets using known and extracted opinion words and targets through the identification of syntactic relations. The identification of the relations is the key to the extractions. The method is evaluated on a product review dataset, and the results show that the approach outperforms existing methods in both opinion lexicon expansion and target extraction. The method is also effective in assigning sentiment polarities to newly extracted opinion words and pruning noisy targets. The results demonstrate that the proposed method is effective in extracting a large number of new opinion words and targets with high precision and recall. The method is also effective in handling different numbers of seeds, with the best performance achieved when using a moderate number of seeds. The method is also effective in handling different types of opinionated expressions, including those with negations and contraries. The method is also effective in handling different types of product reviews, including those with complex syntactic structures. The method is also effective in handling different types of opinionated expressions, including those with negations and contraries. The method is also effective in handling different types of product reviews, including those with complex syntactic structures.