Website-based scoring method for Chinese news information multi-dimensional classification

Abstract

The invention discloses a website-based scoring method for Chinese news information multi-dimensional classification. The method comprises the steps of defining five classification levels for Chinese news information: defining different numbers for all categories in the step A; and correspondingly defining different background basic scores for all the categories with the numbers, wherein a calculation formula for a total score is as follows: the total score = the background basic score of the first classification level * the background basic score of the second classification level * the background basic score of the third classification level * the background basic score of the fourth classification level * the background basic score of the fifth classification level. According to the method, the news information is subjected to refined classification from multiple dimensions; the classification is clear, so that subsequent search and deep analysis of the news information are facilitated; after news is scored, all scores are averaged to obtain an average value as importance of the news; the scores can be divided into positive and negative (+-) scores according to emotions, and an absolute value after addition of the positive and negative scores is a social positive and negative energy influence score; and a result obtained after addition and subtraction of the score absolute value and the social positive and negative energy influence score is argumentativeness of the news.

Description

Technical field [0001] The invention relates to a website-based information classification method, in particular to a scoring method for multi-dimensional classification of Chinese news information based on a website. Background technique [0002] News is "a report of recent facts." News uses concise text to summarize a wealth of information and is frequently updated, and through the era of public media dissemination, the source of news is enriched and the dissemination of news is accelerated. However, in the face of explosive growth and disorderly news, it is more difficult for users to obtain the required information. Therefore, there is an urgent need for effective information organization of news. [0003] On the morning of April 19, 2017, a seminar on cyber security and informatization was held in Beijing, emphasizing that promoting my country’s economic and social development in accordance with the development concepts of innovation, coordination, green, openness, and sharin...

AI Technical Summary

Problems solved by technology

[0007] 2. News noise affects the quality of news classification: the authenticity of online news is insufficient, often news titles and texts are inconsistent, expressions before and after the text are inconsistent, headlines are entertaining and kitsch, and exaggerated reports

Text classification of distorted news will obviously interfere with the text representation link, which will cause the classification results to have no practical significance. It is too exaggerated, so that the emotional tendency of the emotional feature words is exaggerated, which in turn affects the results of emotional classification

[0008] 3. The classification system is too simple, which is not conducive to in-depth analysis: the classification system in the current network news classification research is too simple, and the method of artificially selecting categories is mostly adopted. The selected classification system has few categories, few levels, and a degree of differentiation between categories. large, idealized

[0009] 4. The classification dimension is too single: the classification of current online news is mostly carried out from the subject dimension

At present, there have been researches on topic tracking from the time dimension, emotional tendency analysis from the emotional dimension, and classification from the geographical location dimension, but there are still very few studies integrating multiple dimensions. This is a research direction in the future.

[0010] 5. News topics are flat and lack depth: users’ comprehensive cognitive needs for topics or events promote the development of online news topics

News topics should be a kind of in-depth report, but currently many topics are not of high quality, usually just simple listing and accumulation of relevant information. Although the integration of information is realized, the hierarchical relationship between information is ignored, lacking in order and logic. It also lacks systematic sorting, induction and summary, giving users a redundant and messy feeling

[0011] 6. No combined analysis: a piece of news does not necessarily only involve a single aspect, but may involve many aspects

Images