Donation crowd funding field-oriented recommendation method

Abstract

The invention discloses a donation crowd funding field-oriented recommendation method, which is characterized by comprising the following steps of 1, obtaining a data source through a web crawler, andperforming local storage; 2, performing data preprocessing on the acquired data to form standard data; 3, user intention donation crowd funding data association: constructing a donation crowd fundingknowledge base through a natural language method; carrying out information extraction on the text content of the user based on an entity information extraction technology; secondly, storing the extracted association relationship between the user information and the public welfare information into a knowledge base; and finally, knowledge retrieval and reasoning are performed on the knowledge baseto complete analysis of two donation intention association relationships. 4, based on a knowledge base, the invention discloses a recommendation method for constructing user donation crowd funding. Obtaining a donation crowd funding association relationship according to the established donation crowd funding association relationship; a collaborative filtering algorithm is used for calculating theintention donation degree between intentional donators, an intention donation degree matrix is established, Sim (Vp, Vq) represents the intention donation degree between stars with intention donation,Vp and Vn represent users and commodities, and rvp and rvq are matrixes of the two users respectively.

Description

technical field [0001] The invention relates to a recommendation system method based on network donation and crowdfunding information. This paper adopts the combination of data mining, natural language processing and recommendation system methods. Background technique [0002] With the development of the Internet, the characteristics of Internet data have become increasingly prominent, and its data volume has grown extremely rapidly, with various types of data, uneven data quality, and complex correlations. At the same time, another outstanding feature of Internet big data is its low value density. Big data contains a lot of repetition, noise and garbage data, and there are a lot of co-occurring but meaningless correlation patterns. The huge scale and complex relationship of data make the traditional text analysis and mining technology surge in the computational complexity of time and space; in addition, the rapid data growth rate and huge data volume also make the tradition...

AI Technical Summary

Problems solved by technology

However, there are many types of crowdfunding information, and the recommendation range is relatively wide. For users who do not have a similar feature or browsing history, it is currently difficult to push accurate information when users experience a cold start.

Images