Authoritative author and high-quality paper recommending system and recommending method

Abstract

The invention discloses an authoritative author and high-quality paper recommending system and recommending method. According to each preset theme, the recommending system calculates an author authoritative value and a paper quality paper through a plurality of factors including an author level, a citation rate, publish date and a publish periodical or a meeting level of a paper, and hereby recommends an authoritative author and a high-quality paper with the specified theme, so as to avoid authors with low authoritative values or papers with low quality values in an author or paper recommending list, and to reduce system calculation burden and improve system response time. The system and the method, on the basis of properties of academic papers, introduce relates algorithms after taking various factors affecting the author authoritative value and the paper high-quality value into comprehensive consideration, so as to improve accuracy of recommended results, which not only recommends high-quality papers to users but also remarkably shortens calculation time, and the system and the method are good in a real-time updating effect; furthermore, the system can enhance diversity of recommended results in the paper recommending list, and overcome a shortcoming of an existing system which is narrow in user view.

Description

technical field [0001] The present invention relates to a system and method for recommending authoritative authors and high-quality papers. Specifically, it relates to a personalized recommendation system and method for authoritative authors and high-quality papers aimed at user operation behaviors, and belongs to the field of data mining and machine learning. technology field. Background technique [0002] In 2003, D.M.Blei et al. proposed the potential Dirichlet LDA (Latent Dirichlet Allocation) model. The LDA model is based on PLSA (Probabilistic Latent Semantic Analysis), using a K-dimensional hidden random variable γ that obeys the Dirichlet distribution to represent the topic probability distribution of the text, and simulates the text generation process. Later, someone imposed a Dirichlet prior on the η parameter in the model to make it obey the Dirichlet distribution, and the parameter is δ k , which can make the LDA model a complete text generation model. In this...

AI Technical Summary

Problems solved by technology

[0016] (1) Low calculation efficiency: Because the HITS algorithm is a query-related algorithm, real-time calculations must be performed after receiving user queries, and the HITS algorithm itself requires multiple rounds of iterative calculations to obtain the final result, resulting in low calculation efficiency. Low, this is a problem that must be carefully considered in practical applications

[0017] (2) Topic drift problem: If the extended webpage collection contains some pages that are irrelevant to the query topic, and there are many mutual link points between these pages, then using the HITS algorithm is likely to give these irrelevant webpages a high ranking , leading to topic drift in search results, a phenomenon known as the Tightly-Knit Community Effect

[0019] (4) Unstable structure: In the original expanded webpage collection, if individual webpages are added or deleted or a few link relationships are changed, the ranking results of the HITS algorithm will change greatly

The problem with this method is that the calculation efficiency is low, and the system must receive the user's search before it starts to perform calculation operations in real time

For example, since the above method builds a user model based on the papers published by the user, this method cannot build a user model for a user who has not published a paper, and naturally cannot recommend the user's papers

However, as long as the quality of the paper meets the requirements, they will all participate in the calculation of the similarity between the user and the paper, which will inevitably increase the amount of calculation of the system, thereby increasing the waiting time of the user

Images