Search

Patent terminology relation extraction is of great significance to the construction of patent Knowledge graph. In order to solve the problem of long-distance dependency in traditional depth learning, a new method of patent terminology relation extraction is proposed, which combines attention mechanism and bi-directional LSTM model and with keyword strategy. Category keyword features in each sentence obtained by the improved TextRank with the patent text information vectorization added. BiLSTM neural work and attention mechanism are employed to extract the temporal information and sentence-level global feature information. Moreover, pooling layer is added to obtain the local features of the text. Finally, we fuse the global features and local features, and output the final classification results through the softmax classifier. The addition of category keywords improves the distinction of categories. Substantial experimental results demonstrate that the proposed model outperform the state-of-art neural model in patent terminology relation extraction.

The sentiment analysis of film reviews is the basis for obtaining the opinions of the audience, and has a very important influence on the movie public opinion control and the stimulation of potential consumers. This paper focuses on the short-text movie reviews, in which the Chinese emotion feature database and theme feature set are constructed according to the important features in the short-text film review corpus. A sentiment analysis method of short-text movie reviews based on multi-feature fusion (ETAT-LSTM) is proposed to analyze emotional orientation of short-text movie reviews in this paper. This method uses LSTM to obtain the semantic features of short-text movie reviews, and uses attention mechanism to fuse subject features and emotion features. Experimental results show that for emotional classification of movie short-text reviews, our proposed model ETAT-LSTM performance better in precision, call rate and F-value, when compared to basic LSTM, CNN model and the traditional deep learning model based on Attention mechanism.

By analyzing a large number of complaints information, we put forward a method based on rules to extract subject words of urban complaint data. We first segment the title of complaint data using stop words and extract candidate subject words, and then filter them using position features, overlapping word features and dependency features to get subject words. The experimental results show that the method proposed is effective to extract subject words of urban complaint data.

For a given query, searching for entities that conform to the description facts in the given set, in view of this goal, this paper proposes a matching method based on classification and semantic extension. The algorithm firstly to classify the query string into three categories, and extract the key word of different categories of query word. Then the keyword is extended to get the matching word set based on the word2vec word vector model. At last we calculate the score of every entity by the weighted matching method and get results according to the score ranking. After the experiment, the method get the correct rate of 63.2%, which has good applicability, and to a certain extent, it reduces the retrieval failure rate due to the query of the spoken language and diversification.