Sanqi data processing system

[0035] Example 1, reference figure 1 , A system based on notoginseng data analysis,

[0036] Step A: Panax notoginseng data mining;

[0037] Step B: Construction of the Panax Notoginseng database;

[0038] Step C: Add recognition of notoginseng data;

[0039] Step D: Analyze the 37 data to generate a report.

[0040] Embodiment 2, a system based on notoginseng data analysis, and notoginseng data mining includes:

[0041] Step A: Analyze the aggregated quality data of Panax notoginseng to obtain real data;

[0042] Step A two, build a decision tree based on real data;

[0043] Step A three, data normalization: normalize the data in the target data set, mark the original position of the data deleted in the normalization process, and store it in the data buffer set;

[0044] Step A four, result judgment: According to the result of data normalization, select the corresponding decision from the decision tree.

[0045] A system based on Panax notoginseng data analysis. The process of constructing a decision tree based on real data includes: determining the data type to which it belongs according to the element characteristics of the real data, assigning corresponding variables to the nodes on the decision tree according to the data type; establishing the variable The corresponding relationship with the address, the real data corresponding to the data type is stored in the address; the process of assigning corresponding variables to the nodes on the decision tree according to the data type includes: the principle of the largest difference between data type groups will be true The data is segmented, and the real data of each segmentation is assigned a variable and a corresponding node.

[0046] A system based on Panax notoginseng data analysis. Step A1 includes statistical analysis of data; using multiple regression statistical analysis method, the processing process is: divided into three stages: collecting data, analyzing data and performing reasoning; Step A2 , Build a decision tree through the process of recursive segmentation; Decision tree construction: (1) Find the initial split, use the training set as the set to generate the decision tree, and classify the records in the training set; use the attribute domain as the classification index; (2) The decision tree grows to a complete tree, repeating (1) until the records in each leaf node belong to the same category; (3) data pruning to remove noise or abnormal data; (4) association rules.

[0047] Embodiment 3, a system based on notoginseng data analysis, the construction of the notoginseng database includes:

[0048] Step B 1: Obtaining data. Step: Obtaining abstract resources obtained in the data mining of Panax notoginseng;

[0049] Step B 2: Pre-processing step: decompose each piece of information in the summary resource to obtain data information including summary content;

[0050] Step B three: classification processing step: extract and classify the abstract content according to the corresponding information list of notoginseng;

[0051] Step B4: Post-processing step: For each piece of information in the corresponding information list of Panax notoginseng, construct a relevant word semantic database corresponding to the information;

[0052] Step B5: Library construction step: establish an information database, the information database including the corresponding information list and the corresponding document abstract.

[0053] A system based on notoginseng data analysis. The corresponding information list includes: resource information, quality information, circulation information, and knowledge information; resource information includes: origin, variety, scale, output, production and operation objects, and cost; Quality information includes: reference to pharmacopoeial indicators, including grades, specifications, testing indicators, and query information; circulation information includes: processing entities, quantities, specifications; storage locations, methods, and quantities; market entities, distribution, transaction volume, and prices , Annual distribution; pharmaceutical main body, category, quantity, distribution, price; knowledge information includes: policies, regulations, development plans, new research, network evaluation, news reports; the steps to obtain abstract resources include: use the word "sanqi" as the search term , Using web crawler technology to obtain document abstracts related to the thirty-seven names; the pre-processing step further includes: after obtaining the abstract resources, filtering the document abstract resources according to the homophones of the thirty-seven names, and retaining the abstract The content contains a document abstract of the homophone.