Design And Realization Of Commodity Evaluation And Analysis System Based On Emotion Calculation

With the rapid development of e-commerce, the potential consumers can be assisted by fully reading online product reviews before online shopping. At the same time, online shops have accumulated a large amount of product reviews. For the consumer who wants to buy the specified product, he can’t quickly see the whole picture of such many reviews for the product. So, how can the website effectively summarize the product reviews that consumers can get an overview of the product’s advantage and disadvantage in a short time? That is the problem we want to solve.To solve the problem, this thesis proposes a product review analysis system based on effective computing. The thesis first offers an overview of the current state of product review mining related research. Then the system analysis and design is introduced including the objective, architecture design and module design. For the two key modules of the system, based on the combination of association rule algorithm, syntactic parsing, domain dictionary, sentiment dictionary and word similarity algorithm, the thesis proposes the algorithms for product feature mining, polarity word extraction and polarity judgment. Based on the system design and the key algorithms, the system is implemented which contains a flexible product review collection module. Finally, the thesis gives the demonstration and the experiments which show that the system is effective.The proposed product analysis system based on effective computing can accept the product online shopping page link from consumers and then collect the product review raw data automatically by pre-defined configuration. It can extract the product features and the corresponding opinions, more clearly show the summary of the product reviews to consumers, help consumers better understand the product they want to buy. In conclusion, the system can help consumers make better shopping decision, with high practical application value.