Method and device for generating statistical data matrix of user behaviors

Abstract

The invention discloses a method and device for generating a statistical data matrix of user behaviors, and relates to the technical field of computers. One specific embodiment of the method comprisesthe following steps: acquiring a plurality of pieces of single behavior data of a plurality of users for a plurality of articles within a preset time interval; determining a non-zero statistical result of each user for the target dimension data of each article and key value pair data formed by the non-zero statistical result and article identifiers according to the single behavior data; independently storing the key value pair data of the same user in a set, storing the article identifiers in a first subset, and storing the non-zero statistical result in a second subset; and converting the article identifiers in each set into a column serial number or a row serial number of the matrix such that each set becomes a row vector or a column vector for forming a statistical data matrix. According to the embodiment, the statistical result of the user behavior data can be quickly obtained on the premise of not using a structured query language such that the statistical data matrix without wasting the storage space is generated.

Description

technical field [0001] The invention relates to the field of computer technology, in particular to a method and device for generating a statistical data matrix of user behavior. Background technique [0002] In the field of computer technology, it is often faced with a scenario where user behavior needs to be analyzed. The above user behavior may be the user's behaviors such as browsing, searching, and purchasing for items. In order to more intuitively display the behavior discrimination of different users for different items, a statistical data matrix of user behavior can be established. Each row of data in the matrix can correspond to a user, each column of data can correspond to an item, and each element represents a user. Statistical results of behavioral data for an item within a preset time interval. [0003] In the existing distributed systems using Hive (a data warehouse tool applied in distributed systems) as the query engine, the above statistics are generally obt...

AI Technical Summary

Problems solved by technology

In actual scenarios, the number of item types is often relatively large, which makes the writing, storage, and execution of Hive SQL statements consume a lot of resources, which seriously affects the generation efficiency of the above statistical data matrix

At the same time, the number of users is often also relatively large, and the above-mentioned statistical data matrix is ​​generally a sparse matrix, and the statistical data matrix stored in the existing method (that is, the method of displaying each element of the matrix) is slow to read and wastes a lot of space

Images