【摘要】 用户在微博等社交网络上随时发布自己身边的事件,这些大量的事件都有可能成为新闻的线索。与此同时,新闻工作者在面对海量的信息数据时,仍然依靠新闻记者个人的经验直觉,或者社会公众主动提供有关信息等传统手段去发掘新闻线索,往往会显得力不从心。本文在现有理论基础和新闻专业人员的实践基础上,构建了新闻价值评价的模型,然后构建了利用自动化技术实现新闻线索发现的总体框架。本文通过对微博用户创作内容(UGC)进行信息抽取和处理,并将抽取的相关信息实体分为多个变量维度,根据建立的新闻线索发现模型进行计算,从中找到有价值的新闻线索,将结果通过新闻线索发现系统进行数据可视化展示。用最直观的方式帮助媒体随时获知网络中发布报道的价值指数,缩短新闻生产链长度,提高时效性,在竞争中获取更大的优势。更多还原

【Abstract】 People tend to publish events that happened around them.In this era of information explosion all these data are likely to become clues of news.However,the number of Internet users is quite large,even a single micro-blog platform has tens of millions of active users every day.Traditional means as personal intuition or information provided by the public to found out news clues in such huge amounts of information are not reliable to the journalists any more.Thus,data mining to news clues based on big data has become an important topic in news industry.Data collected from Weibo user-generated content has a series of preprocessing steps including data normalization,outlier cleaning and redundant removal for a more effective lexical analysis result.Then a news clues discovery system is designed and implemented.Features involve eventtime,event result,sensitive words and named entities are extracted and calculated according to the news clues discovery model to found out valuable clues.The result is shown by a data visualization module in the system.更多还原