Big Data Acquisition And Application Ofproduct Reviews Based On Electronic Business Platform

With the rapid development of electronic commerce, online shopping is becoming more and more frequent, each electronic business platform will generate a lot of daily trading data and user reviews data, there are a lot of valuable information about the user’s comments, such as the product’s defect information, user’s demand information, etc.. In this regard, this paper deeply studies the acquisition and application of product review data in electronic business platform, by extracting the evaluation opinions and opinions of each attribute word in the big data of product reviews, after the extraction of the point of view to generate an evaluation summary, to form a concise and readable presentation to the user, provide a better reference and guidance for customers shopping, and enable businesses to better understanding of customer needs, improve service quality, enable product designers to understand the user’s experience in a timely manner, improve product design deficiencies, improve product quality.In this regard, this paper makes a deep research on the acquisition, clustering, product attribute word recognition, and the extraction and integration of the product reviews, the Nutch web crawler is combined with Hadoop to realize the distributed crawling of the review data, which improves the efficiency of the review data; the review data will be taken from the data set, TF-IDF method is used to calculate the weights of the feature words, based on the vector space model, the similarity of the comment sentences is calculated, and the Canopy clustering algorithm and K-means clustering algorithm used in combination, using MapReduce framework to achieve the two algorithms on the review of data clustering analysis, improved the clustering efficiency and accuracy of the data.After clustering analysis of the massive review data, the product attributes are obtained as the main cluster centers, an evaluation method based on product attributes is adopted, the evaluation of product attributes in each cluster group is extracted, after the view of the formation of a summary of the evaluation summary, presenting to the user in the form of evaluation summary, which improves the reading ability of the opinion of attribute word evaluation. In order to standardize the extraction of attribute words views in comment statement, this paper constructs the reviews of data quality assessment system, to ensure the quality of the extracted attributes word view, to provide a reference for the extraction algorithm of the optimization.Finally, use XX water purifier as an example, through cluster analysis, attribute words opinion extraction and integration analysis, generated XX water purifier evaluation abstract, the advantages and disadvantages of XX water purifier are obtained, and a brief analysis is made on the application of the evaluation of the extracted attribute words. By the application test show that this review data acquisition and analysis method is correct and effective. The evaluation of the extracted attributes is great significance to the designers and users, providing new ideas and methods for the acquisition, analysis and application of big data in the electronic business platform.