With the rise of mobile social networks, an increasing number of consumers are shopping through Internet platforms. The information asymmetry between consumers and producers has caused producers to misjudge the positioning of agricultural products in the market and damaged the interests of consumers. This imbalance between supply and demand is detrimental to the development of the agricultural market. Sentiment tendency analysis of after-sale reviews of agricultural products on the Internet could effectively help consumers evaluate the quality of agricultural products and help enterprises optimize and upgrade their products. Targeting problems such as non-standard expressions and sparse features in agricultural product reviews, this paper proposes a sentiment analysis algorithm based on an improved Bidirectional Encoder Representations from Transformers (BERT) model with symmetrical structure to obtain sentence-level feature vectors of agricultural product evaluations containing complete semantic information. Specifically, we propose a recognition method based on speech rules to identify the emotional tendencies of consumers when evaluating agricultural products and extract consumer demand for agricultural product attributes from online reviews. Our results showed that the F1 value of the trained model reached 89.86% on the test set, which is an increase of 7.05 compared with that of the original BERT model. The agricultural evaluation classification algorithm proposed in this paper could efficiently determine the emotion expressed by the text, which helps to further analyze network evaluation data, extract effective information, and realize the visualization of emotion.

This study designed a handheld vibrating coffee harvester to improve the mechanized harvesting of Coffea arabica L. The proposed device was used to vibrate branches of Coffea arabica L. trees, and the shedding of coffee fruit and the operation parameters of vibrational harvesting were analyzed. Images captured using high-speed photography were used to derive a force equation that represents the forced vibration of the coffee fruit-stalk joint. In addition, the vibrations of coffee berries and branches were theoretically analyzed, and the results were used to establish a dynamic vibration model of coffee trees. The shedding of coffee berries was primarily affected by the vibration frequency, vibration amplitude, and excitation position, which were simulated using a rigid–flexible branch-machine coupling model on RecurDyn software. Furthermore, field experiments were conducted to determine the optimal working parameters for coffee harvesting using vibrations. The results indicated optimal picking performance when the vibration frequency, vibration amplitude, and excitation position were 62 Hz, 9 mm, and 0.4 L, respectively. The harvesting rates of ripe and unripe coffee were 92.22% and 8.33%, respectively, and the damage rate was 5.23%. Thus, the proposed harvester can satisfactorily achieve the optimal harvesting of Coffea arabica L.