技术领域technical field
本发明属于计算机应用领域，涉及将自然语言处理技术、本体理论和语义关联技术应用于互联网舆情突发事件的智能识别和防控方案的自动生成。它基于计算机信息处理方法实现对应急防控预案的格式化转化，实现舆情突发事件情景与预案之间的语义匹配，实现对各种互联网舆情突发事件的准确识别和辅助决策。The invention belongs to the field of computer applications, and relates to the application of natural language processing technology, ontology theory and semantic association technology to intelligent identification of Internet public opinion emergencies and automatic generation of prevention and control schemes. Based on computer information processing methods, it realizes the format transformation of emergency prevention and control plans, realizes the semantic matching between public opinion emergency scenarios and plans, and realizes accurate identification and auxiliary decision-making of various Internet public opinion emergencies.
背景技术Background technique
随着互联网技术的不断发展，互联网已成为一种被广泛使用的大众媒介，其触角几乎伸向社会的各个领域，并逐渐成为公众舆论的一个新的重要媒介。网络舆情是公众在互联网上公开表达的对某种社会现象或社会问题的具有一定影响力和倾向性的共同意见，网络舆情对政治生活秩序和社会稳定的影响与日俱增,一些网络舆情突发事件不能及时妥善处理,极有可能诱发民众的不良情绪及不良行为的发生,进而对社会稳定形成严重威胁。迫切需要一种技术手段能够实现对网络舆情信息的自动监控，能够对舆情突发事件的处置提供决策支持。With the continuous development of Internet technology, the Internet has become a widely used mass media, its tentacles extend to almost every field of society, and gradually become a new important media of public opinion. Internet public opinion is a common opinion with certain influence and tendency expressed by the public on the Internet on a certain social phenomenon or social problem. The influence of Internet public opinion on the order of political life and social stability is increasing day by day. Some Internet public opinion emergencies cannot Timely and proper handling, it is very likely to induce people's bad emotions and bad behaviors, and then pose a serious threat to social stability. There is an urgent need for a technical means that can realize the automatic monitoring of network public opinion information and provide decision support for the handling of public opinion emergencies.
发明内容Contents of the invention
本发明就是针对上述需求，提出了一种计算机应用系统―舆情突发事件应急处理系统，它能够对互联网舆情进行实时监控，能够辅助决策者根据舆情突发事件的实际情况有针对性地形成相适合的防控方案，加快对网络舆情突发事件的处置响应速度。In response to the above needs, the present invention proposes a computer application system-public opinion emergency response system, which can monitor Internet public opinion in real time, and can assist decision makers to form relevant information in a targeted manner according to the actual situation of public opinion emergencies. Appropriate prevention and control programs can speed up the response to online public opinion emergencies.
本发明所要解决的技术问题由以下技术方案实现：The technical problem to be solved by the present invention is realized by the following technical solutions:
一种智能的舆情突发事件应急处理系统，其特征在于：该系统包括互联网信息采集与解析模块、互联网信息分析模块、网络文本类别判断与聚类分析模块、应急处理方案生成模块和应急处理效果评估模块；所述互联网信息采集与解析模块用于从互联网上采集信息，抽取出网页中自然语言文字以及网页的元数据信息，并保存到数据库中；所述互联网信息分析模块用于对采集来的信息中的自然语言文字进行特征抽取，形成文本特征；所述网络文本类别判断与聚类分析模块用于对网络文本的类别进行判断，对累积网络文本进行聚类分析；所述应急处理方案生成模块用于根据舆情事件的具体情况自动生成相应的处理预案，决策人员可以基于处理预案制定执行方案；所述应急处理效果评估模块用于对执行方案的执行效果进行评估。An intelligent emergency response system for public opinion emergencies, characterized in that the system includes an Internet information collection and analysis module, an Internet information analysis module, a network text category judgment and cluster analysis module, an emergency treatment plan generation module and an emergency treatment effect Evaluation module; the Internet information collection and analysis module is used to collect information from the Internet, extract the natural language text and metadata information of the webpage in the webpage, and save it in the database; the Internet information analysis module is used to collect information The natural language characters in the information are extracted to form text features; the network text category judgment and cluster analysis module is used to judge the network text category and perform cluster analysis on the accumulated network text; the emergency treatment plan The generation module is used to automatically generate corresponding treatment plans according to the specific conditions of public opinion events, and decision-makers can formulate execution plans based on the treatment plans; the emergency treatment effect evaluation module is used to evaluate the execution effect of the execution plan.
一种智能的舆情突发事件应急处理系统及方法，其特征在于该方法包括以下步骤:An intelligent public opinion emergency emergency handling system and method, is characterized in that the method comprises the following steps:
①互联网信息采集与解析：由连接互联网的计算机从互联网论坛、博客、新闻网站上采集论坛帖子、博客内容和网站新闻网页等网络数据，然后，利用计算机采用基于规则的信息抽取技术自动地对网络数据进行解析，从其中抽取两类信息：自然语言文字信息和网页的元数据信息；自然语言文字信息包括新闻标题、新闻正文、论坛帖子标题、帖子内容等信息；网页的元数据信息包括发表时间、作者、发帖者、帖子回复量、帖子阅读量、出现的网站名称、网站URL等信息，解析出来的信息保存到数据库中，信息采集与解析是一个持续的过程，形成对互联网站的自动连续监控；①Internet information collection and analysis: The computer connected to the Internet collects network data such as forum posts, blog content, and website news pages from Internet forums, blogs, and news websites, and then uses the computer to automatically analyze the information on the Internet using rule-based information extraction technology. Analyze the data and extract two types of information from it: natural language text information and web page metadata information; natural language text information includes news titles, news texts, forum post titles, post content and other information; web page metadata information includes publication time , author, poster, post reply volume, post reading volume, website name, website URL and other information, the parsed information is stored in the database. Information collection and analysis is a continuous process, forming an automatic continuous process of Internet sites monitor;
②互联网信息分析：首先利用自然语言处理技术的中文分词方法对网络文本的标题和正文内容分别进行分词，并对分词结果中每个词项的词性进行标注，之后舍弃掉除名词、动词、形容词之外的词项，然后利用文本多精度表示方法抽取网络文本的单个词项特征和词项关联特征，再根据分词结果中的词性标注情况识别出网络文本中的地理位置特征和人物特征，地理位置特征是网络文本中出现的地理位置名称、人物特征是网络文本中出现的人物名称；②Internet information analysis: First, use the Chinese word segmentation method of natural language processing technology to segment the title and text content of the network text, and mark the part of speech of each word item in the word segmentation results, and then discard nouns, verbs, and adjectives. Then use the text multi-precision representation method to extract the features of individual words and word association features of the network text, and then identify the geographical location features and character features in the network text according to the part-of-speech tagging in the word segmentation results. The location feature is the geographical location name appearing in the network text, and the character feature is the person name appearing in the network text;
③将步骤②处理后的网络文本中的词项与计算机数据库中设定的舆情类别的词项特征进行比对匹配，并根据匹配结果将网络文本按照计算机数据库中设定的舆情类别进行归类处理；将不能归类的网络文本进行聚类分析，把内容相近的网络文本聚成簇，若簇内网络文本数量超出设定阈值，则对簇内网络文本进行舆情类别的词项特征抽处理，并将抽取的舆情类别的词项特征补充到计算机数据库中；对于完成归类的网络文本转入步骤④；其中，匹配内容包括单个词项特征、词项关联特征、地理位置特征和人物特征；③ Compare and match the words in the network text processed in step ② with the word features of the public opinion category set in the computer database, and classify the network text according to the public opinion category set in the computer database according to the matching results Processing: cluster analysis of network texts that cannot be classified, and cluster network texts with similar content. If the number of network texts in a cluster exceeds the set threshold, the network texts in the cluster will be processed by word feature extraction of public opinion categories , and add the extracted word features of the public opinion category to the computer database; for the network texts that have been classified, turn to step ④; where the matching content includes single word features, word association features, geographical location features and character features ;
④如果在指定时间段内，属于某一类别的网络文本的数量或者出现该类别网络文本的网站数量超过指定的阈值，则启动应急预案；④ If within a specified period of time, the number of web texts belonging to a certain category or the number of websites that appear in this category of web texts exceeds the specified threshold, activate the emergency plan;
完成智能舆情突发事件的应急处理。Complete the emergency handling of intelligent public opinion emergencies.
其中，在步骤④之后还包括应急处理效果评估步骤：首先按照评估指标采集指标数据，然后将指标数据输入评估公式得出量化评估结果。Among them, after step ④, there is also a step of evaluating the effect of emergency response: firstly collect index data according to the evaluation index, and then input the index data into the evaluation formula to obtain a quantitative evaluation result.
其中，在步骤③中根据匹配结果将网络文本按照计算机数据库中设定的舆情类别进行归类处理具体为：网络文本类别判断的方法是将网络文本的词项与每个舆情类别的词项特征进行比对匹配，分别在单个词特征、词关联特征、地理位置特征和人物特征四个方面进行匹配运算，根据匹配情况得到网络文本与各个舆情类别的相似度值，将文本归属为相似度值最高的舆情类别。Among them, in step ③, according to the matching results, the network texts are classified according to the public opinion categories set in the computer database. Perform comparison and matching, and perform matching operations in four aspects: single word features, word association features, geographical location features, and character features, and obtain the similarity value between the network text and each public opinion category according to the matching situation, and attribute the text to the similarity value The highest sentiment category.
其中，步骤③中对簇内网络文本进行舆情类别的词项特征抽处理，具体为：假设簇T包含的网络文本有T＝{t1,t2,…tn},利用文本多精度表示方法抽取出每个文本ti的单个词项特征和词项关联特征，再采用统计方法计算出T中所有文本的所有单个词项特征和词项关联特征的统计分布规律，选择在T中一半以上网络文本中出现过的词汇作为舆情类别词项特征，并计算出其在T内的平均发生频率作为舆情类别特征词项的频率；其中，1≤i≤n。Among them, in step ③, the network text in the cluster is subjected to feature extraction processing of public opinion categories, specifically: assuming that the network text contained in the cluster T has T={t1 ,t2 ,...tn }, using text multi-precision representation The method extracts the single term features and term association features of each text ti , and then uses the statistical method to calculate the statistical distribution law of all single term features and term association features of all texts in T, and selects half of T in T The words that have appeared in the above network texts are used as the characteristics of public opinion category terms, and their average occurrence frequency in T is calculated as the frequency of public opinion category feature terms; where 1≤i≤n.
其中，步骤④中应急预案的生成方法为：基于互联网舆情事件情景本体知识库模型和网络舆情防控措施预案本体知识库，利用语义匹配技术根据舆情事件情景的具体情况，从防控措施预案库中自动匹配出最适合的应急处理预案。Among them, the generation method of the emergency plan in step ④ is: based on the knowledge base model of the Internet public opinion event scenario ontology and the ontology knowledge base of the network public opinion prevention and control measures, using semantic matching technology according to the specific situation of the public opinion event scenario, from the prevention and control measures pre-plan database automatically match the most suitable emergency response plan.
与现有的技术相比，本发明具有以下的优点和有益效果:Compared with prior art, the present invention has following advantage and beneficial effect:
1、本发明不仅能够对网络舆情进行自动监控，还能够针对突发舆情事件给出防控措施方案。1. The present invention can not only automatically monitor network public opinion, but also provide prevention and control measures for sudden public opinion events.
2、本发明的舆情类型识别计算机数据库具有可扩展性，通过文本聚类分析不断补充新型舆情类型特征到数据库中，使系统能够识别新增类型的舆情事件。2. The public opinion type identification computer database of the present invention has scalability, and new public opinion type features are constantly added to the database through text cluster analysis, so that the system can identify newly added types of public opinion events.
附图说明Description of drawings
图1系统模块组成图Figure 1 System module composition diagram
图2舆情分类体系模型图Figure 2 Model Diagram of Public Opinion Classification System
图3舆情分类体系概念属性模型图Figure 3 Conceptual attribute model diagram of public opinion classification system
图4舆情分类体系示意图Figure 4 Schematic Diagram of Public Opinion Classification System
图5类别特征产生过程工作原理图Figure 5 Schematic diagram of the process of category feature generation
图6语义匹配原理图Figure 6 Schematic diagram of semantic matching
图7基于网络文本聚类的知识扩展图Figure 7 Knowledge expansion diagram based on network text clustering
图8舆情事件情景本体知识库图Figure 8 Knowledge Base Map of Public Opinion Event Scenario Ontology
图9舆情防控措施预案本体知识库图Figure 9 Ontology knowledge base map of public opinion prevention and control measures and plans
图10网络舆情防控知识语义模型图Figure 10 Semantic Model of Internet Public Opinion Prevention and Control Knowledge
图11基于语义的匹配方法图Figure 11 Semantic-based matching method diagram
图12应急处理效果评估指标体系图Figure 12 Evaluation Index System Diagram of Emergency Treatment Effect
具体实施方式Detailed ways
下面将结合附图和具体实施例对本发明做进一步说明。但本发明的实施方式不限于此。The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments. However, the embodiments of the present invention are not limited thereto.
本实施例提供一种智能的舆情突发事件应急处理系统，该系统包括互联网信息采集与解析模块，互联网信息分析模块，网络文本类别判断与聚类分析模块，应急处理方案生成模块，应急处理效果评估模块，如附图1所示；所述互联网信息采集与解析模块用于从互联网上采集信息，抽取出网页中自然语言文字以及网页的元数据信息，并保存到数据库中；所述互联网信息分析模块用于对采集来的信息中的自然语言文字进行特征抽取，形成文本特征；所述网络文本类别判断与聚类分析模块用于对网络文本的类别进行判断，对累积网络文本进行聚类分析；所述应急处理方案生成模块用于根据舆情事件的具体情况自动生成相应的处理预案，决策人员可以基于处理预案制定执行方案；所述应急处理效果评估模块用于对执行方案的执行效果进行评估。The present embodiment provides an intelligent public opinion emergency response system, which includes an Internet information collection and analysis module, an Internet information analysis module, a network text category judgment and cluster analysis module, an emergency treatment plan generation module, and an emergency treatment effect Evaluation module, as shown in accompanying drawing 1; Described Internet information collecting and parsing module is used for collecting information from Internet, extracts the metadata information of natural language text and webpage in the webpage, and saves in the database; Said Internet information The analysis module is used to extract the features of the natural language characters in the collected information to form text features; the network text category judgment and cluster analysis module is used to judge the network text category and cluster the accumulated network text Analysis; the emergency response plan generating module is used to automatically generate corresponding processing plans according to the specific circumstances of the public opinion event, and decision-makers can formulate an execution plan based on the processing plan; the emergency response effect evaluation module is used to evaluate the execution effect of the execution plan Evaluate.
本实施例还提供一种智能的舆情突发事件应急处理系统的工作方法，该方法包括以下步骤:The present embodiment also provides a working method of an intelligent public opinion emergency emergency handling system, the method comprising the following steps:
①互联网信息采集与解析：由连接互联网的计算机从互联网论坛、博客、新闻网站上采集论坛帖子、博客内容和网站新闻网页等网络数据，然后，利用计算机采用基于规则的信息抽取技术自动地对网络数据进行解析，从其中抽取两类信息：自然语言文字信息和网页的元数据信息。自然语言文字信息包括新闻标题、新闻正文、论坛帖子标题、帖子内容、作者、发帖者等信息；网页的元数据信息包括发表时间、帖子回复量、帖子阅读量、出现的网站名称、网站URL等，解析出来的关键信息保存到数据库中，信息采集与解析是一个持续的过程，形成对互联网站的自动连续监控。①Internet information collection and analysis: The computer connected to the Internet collects network data such as forum posts, blog content, and website news pages from Internet forums, blogs, and news websites, and then uses the computer to automatically analyze the information on the Internet using rule-based information extraction technology. The data is analyzed, and two types of information are extracted from it: natural language text information and metadata information of web pages. Natural language text information includes news titles, news texts, forum post titles, post content, authors, posters, and other information; web page metadata information includes publishing time, post replies, post readings, website names that appear, website URLs, etc. , the analyzed key information is stored in the database, and information collection and analysis is a continuous process, forming an automatic continuous monitoring of Internet sites.
②互联网信息分析：首先利用自然语言处理技术的中文分词方法对网络文本的标题和正文内容进行分词和词性标注处理，标注出每个词项的词性，舍弃掉文本中除名词、动词、形容词之外的词汇。然后利用已获得国家发明专利授权的“一种用于文本检索系统的文本多精度表示方法”所述方法抽取网络文本的单个词特征和词关联特征。另外，根据分词结果中的词性标注情况识别出文本中的地理位置特征和人物特征，地理位置特征是网络文本中出现的地理位置名称、人物特征是网络文本中出现的人物名称，如附图5中的网络文本语义特征抽取功能单元所示。总的来说网络文本的特征是一组词汇，配有其发生频率。②Internet information analysis: Firstly, use the Chinese word segmentation method of natural language processing technology to perform word segmentation and part-of-speech tagging on the title and text content of the network text, mark the part of speech of each word item, and discard all words except nouns, verbs, and adjectives in the text. foreign vocabulary. Then use the method described in "A Text Multi-precision Representation Method for Text Retrieval System" that has been authorized by the national invention patent to extract single word features and word association features of network texts. In addition, according to the part-of-speech tagging in the word segmentation results, the geographic location features and character features in the text are identified. The geographic location features are the geographic location names that appear in the network text, and the character features are the person names that appear in the network text, as shown in Figure 5 The functional unit of network text semantic feature extraction is shown in . In general, web texts are characterized by a set of words, accompanied by their frequency of occurrence.
③网络文本类别判断与聚类分析：其目的是基于网络文本的内容采用文本分类技术判断网络文本的所属类别。所属类别是基于本体论事先建立起来的一个舆情分类体系模型中的一种，舆情分类体系模型如附图2所示，它是一个层次化模型，第一层是大类，第二层是小类，每个小类都由概念属性来定义，如附图3所示，有两个概念属性：类别语义特征和防控策略。类别语义特征包括：③ Network text category judgment and cluster analysis: its purpose is to use text classification technology to judge the category of network text based on the content of network text. The category to which it belongs is one of the public opinion classification system models established in advance based on ontology. The public opinion classification system model is shown in Figure 2. It is a hierarchical model. Each sub-category is defined by conceptual attributes, as shown in Figure 3, there are two conceptual attributes: category semantic features and prevention and control strategies. Class semantic features include:
单个词特征：类别语义特征抽取模块抽取出的网络文本的单个词特征；Single word feature: the single word feature of the network text extracted by the category semantic feature extraction module;
词关联特征：类别语义特征抽取模块抽取出的网络文本的多词关联特征；Word association feature: the multi-word association feature of the network text extracted by the category semantic feature extraction module;
地理位置特征：类别语义特征抽取模块抽取出的网络文本中的地理位置名称；Geographic location feature: the geographical location name in the network text extracted by the category semantic feature extraction module;
人物特征：类别语义特征抽取模块抽取出的网络文本中的人物名称；Character features: the names of characters in the network text extracted by the category semantic feature extraction module;
实例：该类型网络舆情的一个实例文本；Example: an example text of this type of Internet public opinion;
类别判断准则。判断一批某类舆情相关的文本累积是否真的是一次舆情事件。例如，IF出现舆情文本的网站数量大于n THEN是一次舆情事件；IF舆情文本的回帖数量大于n THEN是一次舆情事件。category judgment criteria. Judging whether the accumulation of a batch of texts related to a certain type of public opinion is really a public opinion event. For example, if the number of websites with IF public opinion texts is greater than n THEN, it is a public opinion event; if the number of replies to IF public opinion texts is greater than n THEN, it is a public opinion event.
防控策略包括防控原则和防控方法，防控原则是针对某类舆情事件开展防御和控制的基本原则；防控方法是针对某类舆情采取的具体防控措施。The prevention and control strategy includes prevention and control principles and prevention and control methods. The prevention and control principles are the basic principles for the defense and control of a certain type of public opinion events; the prevention and control methods are specific prevention and control measures for a certain type of public opinion.
图4是一个实际舆情分类体系的示意图。Figure 4 is a schematic diagram of an actual public opinion classification system.
每一个类别都有其类别特征，为每个类别产生类别特征的方法如附图5所示：首先采集各个类别的若干网络文本作为训练样本，利用自然语言处理技术的中文分词方法对所有训练样本进行分词和词性标注处理，标注出每个词项的词性，舍弃掉文本中除名词、动词、形容词之外的词汇；由网络文本语义特征抽取功能单元抽取每个文本的单个词特征、词关联特征、地理位置特征和人物特征，再由类别语义特征抽取功能单元抽取类别语义特征；具体方法是：利用计算机采用统计算法计算出每个文本的各个特征在每个类别内以及训练样本全集的统计分布规律，选择在一半以上类别样本文档中出现过且不是训练样本全集内所有样本所共有的词汇作为类别特征词，并计算出其类别内平均发生频率作为类别特征词的频率。总的来说类别特征是一组代表类别特征的词汇，配有其平均发生频率。Each category has its category characteristics, and the method for generating category characteristics for each category is shown in Figure 5: first collect several network texts of each category as training samples, and use the Chinese word segmentation method of natural language processing technology to analyze all training samples Carry out word segmentation and part-of-speech tagging, mark the part of speech of each lexical item, and discard words other than nouns, verbs, and adjectives in the text; the network text semantic feature extraction function unit extracts the individual word features and word associations of each text feature, geographic location feature and character feature, and then the category semantic feature is extracted by the category semantic feature extraction functional unit; the specific method is: use the computer to calculate the statistics of each feature of each text in each category and the complete set of training samples by using statistical algorithms. Distribution law, select the words that appear in more than half of the sample documents of the category and are not common to all samples in the training sample set as the category feature words, and calculate the average occurrence frequency in the category as the category feature words. In general, categorical features are a set of words representing categorical features, with their average frequency of occurrence.
网络文本类别判断的方法是将网络文本的特征词项与每一个类别特征词项进行比对匹配，如附图6所示，分别在单个词特征、词关联特征、地理位置特征和人物特征四个方面进行匹配运算，并按照下面的公式计算相似度值，将文本归属为相似度值最高的类别。The method of network text category judgment is to compare and match the feature terms of the network text with each category feature term. According to the matching operation, the similarity value is calculated according to the following formula, and the text is assigned to the category with the highest similarity value.
其中，in,
d表示待分类文档；d represents the document to be classified;
C表示类别；C means category;
coord(d,C)表示待识别文本d中包含类别C的类别特征词项的数量；coord(d,C) indicates the number of category feature terms containing category C in the text d to be recognized;
frequency表示特征词项t在类别特征中的词频； frequency indicates the word frequency of the feature term t in the category feature;
weight(t)：表示特征词项t的权重；weight(t): Indicates the weight of the feature term t;
frequency和weight值可以从建模过程中创建的类别特征词项表中获得，类别特征词项表如表1所示。The frequency and weight values can be obtained from the category feature vocabulary item table created during the modeling process, and the category feature vocabulary item table is shown in Table 1.
表1 类别特征词项表Table 1 Class feature vocabulary item table
numofClasses：表示共有几个类别；numofClasses: Indicates how many categories there are;
ClassFreq(t)：表示特征项项t同时是几个类别的特征词项。ClassFreq(t): Indicates that the feature item t is a feature term of several categories at the same time.
如附图7所示，网络文本经过预处理功能单元处理后，获得文本分词结果并去除停用词，再通过语义特征抽取模块得到其语义特征，利用网络文本类别判断功能单元判读其是否为已知的n种网络舆情的一种，若是则将其归类，否则，将其转给网络文本聚类分析功能单元进行分析，看其中是否有热点话题,对采集来到每一个网络文本都进行类别判断，符合分类条件的网络文本被赋以相应的类别标签。如果在指定时间段内，属于某一类别的网络文本的数量、出现该类别网络文本的网站数量超过指定的阈值，则向系统操作人员发出告警，进而由应急处理方案生成模块给出应急处理方案。As shown in Figure 7, after the network text is processed by the preprocessing functional unit, the word segmentation result of the text is obtained and the stop words are removed, and then its semantic features are obtained through the semantic feature extraction module, and the network text category judgment function unit is used to judge whether it is already If it is one of the known n types of Internet public opinion, it will be classified, otherwise, it will be transferred to the network text clustering analysis function unit for analysis to see if there are hot topics among them, and each network text collected will be analyzed. Category judgment, the network text that meets the classification conditions is assigned the corresponding category label. If within a specified period of time, the number of web texts belonging to a certain category and the number of websites that appear in this category of web texts exceed the specified threshold, an alarm will be sent to the system operator, and then the emergency treatment plan generation module will give an emergency treatment plan .
在上述网络文本类别判断过程中，会出现一些不属于现有舆情分类体系模型中的任何一类的文本，随着时间的推移，未知类型文本会不断累积，对累积的未知类型文本进行聚类分析,把内容相近的网络文本聚成簇，若簇内网络文本数量超出一定阈值，则将其作为热点话题提交人工判读，如果确定其为新的舆情类别，则对其进行舆情类别语义特征抽处理，并将抽取的类别语义特征补充到知识库中，具体过程如附图7所示；上述过程保证了本系统的知识库的可扩展性，使得系统在补充知识后能够识别互联网上的新型舆情。In the above-mentioned network text category judgment process, there will be some texts that do not belong to any category in the existing public opinion classification system model. As time goes by, unknown types of texts will continue to accumulate, and the accumulated unknown types of texts will be clustered Analysis, clustering network texts with similar content into clusters, if the number of network texts in the cluster exceeds a certain threshold, it will be submitted as a hot topic for manual interpretation, if it is determined to be a new public opinion category, the semantic feature extraction of the public opinion category will be carried out process, and add the extracted category semantic features to the knowledge base, the specific process is shown in Figure 7; the above process ensures the scalability of the knowledge base of this system, so that the system can identify new types of information on the Internet after supplementing the knowledge. Public opinion.
④应急处理方案生成：是在舆情类型识别的基础上，针对识别出的舆情类型提供应急处置预案，其特征是，利用本体论技术构建层次化的互联网舆情事件情景本体知识库模型和网络舆情防控措施预案本体知识库模型。前者对舆情事件进行定性和定量的描述，如附图8所示；后者将自然语言文字方式存在的舆情应急防控规章制度、处理规范、应对措施进行数字化，如附图9所示。这样做的目的是将非格式化的信息转变为计算机可理解的格式化信息。有了上述两个知识库模型的支撑，就可以基于计算机利用语义匹配技术自动地实现舆情事件的自动识别，相应防范措施、处理方案的快速自动推理，处理预案的实时辅助生成。情景本体知识库包括舆情、时间、网站、参与者、受众、潜在危害等知识概念。④Generation of emergency response plan: Based on the identification of public opinion types, emergency response plans are provided for the identified types of public opinion. Ontology knowledge base model of control measures plan. The former provides qualitative and quantitative descriptions of public opinion events, as shown in Figure 8; the latter digitizes public opinion emergency prevention and control regulations, processing norms, and countermeasures in natural language, as shown in Figure 9. The purpose of this is to convert unformatted information into formatted information that a computer can understand. With the support of the above two knowledge base models, the automatic recognition of public opinion events, the rapid and automatic reasoning of corresponding preventive measures and processing plans, and the real-time auxiliary generation of processing plans can be automatically realized based on the computer using semantic matching technology. The contextual ontology knowledge base includes knowledge concepts such as public opinion, time, website, participants, audience, and potential hazards.
在互联网信息分析和网络文本类别判断步骤中识别出来的舆情事件的信息会被抽取出来存储到舆情事件情景本体知识库中；舆情类别信息由网络文本类别判断步骤给出，具体采用的是文本分类技术；舆情内容、时间发生时间、时间持续时间、网站名称、网站数量、参与者用户名由互联网信息分析步骤给出的，采用的是基于规则的信息抽取技术；其它信息如舆情等级、参与者IP地址等信息则根据先验知识进行填写。The public opinion event information identified in the Internet information analysis and network text category judgment steps will be extracted and stored in the public opinion event scenario ontology knowledge base; the public opinion category information is given by the network text category judgment step, specifically using text classification Technology; public opinion content, time of occurrence, time duration, website name, number of websites, and participant user names are given by the Internet information analysis step, using rule-based information extraction technology; other information such as public opinion level, participant Information such as the IP address is filled in based on prior knowledge.
舆情防控措施预案本体知识库包括编制依据、适用范围、资源、防控措施四个方面，其内容根据具体的法律法规内容进行填写。The ontology knowledge base of public opinion prevention and control measures and plans includes four aspects: compilation basis, scope of application, resources, and prevention and control measures, and its content is filled in according to specific laws and regulations.
基于互联网舆情事件情景本体知识库和网络舆情防控措施预案本体知识库共同构成了网络舆情防控知识语义模型，基于此模型，利用语义匹配技术生成应急预案，如附图10所示。应急预案是指导处置各种舆情突发事件的方案和方法，而每个舆情事件的具体条件、状况和参数各不相同，决策者需要根据具体情况从防控预案中选定适当的防控处置措施、方法和实施步骤作为应急预案，并调配相应的组织机构和部门执行应急预案。为此，将事件情景的“舆情类别”、“舆情内容”、“舆情等级”分别与预案本体的“适用事件类型”、“适用事件内容”、“适用事件等级”相匹配，如附图11所示，从而发现与舆情事件相适合的应对预案，如表2和表3所示。Based on the Internet public opinion event scenario ontology knowledge base and the network public opinion prevention and control measures ontology knowledge base together constitute the network public opinion prevention and control knowledge semantic model, based on this model, use semantic matching technology to generate emergency plans, as shown in Figure 10. The emergency plan is the plan and method to guide the handling of various public opinion emergencies, and the specific conditions, conditions and parameters of each public opinion event are different, and decision makers need to select appropriate prevention and control measures from the prevention and control plan according to the specific situation The measures, methods and implementation steps are taken as the emergency plan, and the corresponding organizations and departments are deployed to implement the emergency plan. To this end, match the "public opinion category", "public opinion content" and "public opinion level" of the event scenario with the "applicable event type", "applicable event content" and "applicable event level" of the plan body, as shown in Figure 11 As shown in Table 2 and Table 3, we can find the response plan suitable for public opinion events.
表2 基于语义匹配生成的预案示例Table 2 Examples of scenarios generated based on semantic matching
表3 预案示例说明Table 3 Example description of the plan
应对预案只是一个指导性的方案，需要再根据舆情的具体情况，例如，时间、网站、参与者、受众、潜在危害等情况生成一个具体的执行方案。The response plan is only a guiding plan, and a specific implementation plan needs to be generated according to the specific conditions of public opinion, such as time, website, participants, audience, potential hazards, etc.
⑤应急处理效果评估：应急处理效果评估是基于评估指标体系和评估计算公式完成的，评估指标体系包含了需要评估的事项，评估计算公式计算出量化评估结果；评估指标体系如附图12所示，每个指标的详细描述如表4所示。⑤Evaluation of emergency response effects: The evaluation of emergency response effects is completed based on the evaluation index system and evaluation calculation formula. The evaluation index system includes the items that need to be evaluated, and the evaluation calculation formula calculates the quantitative evaluation results; the evaluation index system is shown in Figure 12 , the detailed description of each indicator is shown in Table 4.
表4 应急处理效果评估指标体系Table 4 Emergency treatment effect evaluation index system
舆情强度指标旨在衡量舆情在范围和形式上的情况。①舆情范围指的是舆情的广度，由网站覆盖度、地区覆盖度、网站数量三个指标来衡量。网站覆盖度指的是包含舆情文本的网站占样本网站的比重；样本网站是经过精心选取的，能在一定程度上代表整个网络状态和水平的网站集合；由于各网站的规模级别不同，要对其进行加权处理，出现舆情文本的样本网站越多，说明舆情的范围越广，当实施防控措施后，如果包含舆情文本的网站数量出现减少的趋势说明防控措施发挥了作用。地区覆盖度指的是包含舆情文本的网站的地理分布情况，出现舆情文本的网站分布越广，说明舆情的影响范围越广。网站数量指的是包含舆情文本的网站的总数量，数量越多，说明舆情的影响范围越广。②舆情形式指的是舆情传播的媒介渠道种类、所用网络文本的长短、网络文本的媒体种类。媒介渠道种类可以是BBS、微博、博客、交友平台、电子邮件等，所用的渠道越多，则传播能力越强。所用网络文本的长度越长，则传播能力越强。媒体种类可以是文本、音频、视频，所用媒体种类越多则舆情影响越强。The strength of public opinion indicator is designed to measure the extent and form of public opinion. ①The scope of public opinion refers to the breadth of public opinion, which is measured by three indicators: website coverage, regional coverage, and number of websites. Website coverage refers to the proportion of websites containing public opinion texts in the sample websites; the sample websites are carefully selected and can represent the entire network status and level to a certain extent; It is weighted, and the more sample websites with public opinion texts, the wider the scope of public opinion. After the implementation of prevention and control measures, if the number of websites containing public opinion texts decreases, it means that the prevention and control measures have played a role. Regional coverage refers to the geographical distribution of websites containing public opinion texts. The wider the distribution of websites with public opinion texts, the wider the influence of public opinion. The number of websites refers to the total number of websites containing public opinion texts. The larger the number, the wider the influence of public opinion. ②The form of public opinion refers to the types of media channels for public opinion dissemination, the length of network texts used, and the media types of network texts. The types of media channels can be BBS, Weibo, blogs, dating platforms, emails, etc. The more channels used, the stronger the communication ability. The longer the length of the web text used, the stronger the dissemination ability. The types of media can be text, audio, and video. The more types of media used, the stronger the influence of public opinion.
受众关注度指标旨在反映网络舆情对受众的影响力，通过受众情况、受众响应、受众态度等指标来衡量。①受众情况指的是受舆情影响的受众数量和受众范围，受众数量通过网络文本浏览者IP数量来测量，受众范围通过网络文本浏览者IP的分布地域广度来测量。②受众响应指的是浏览者对网络文本的关注程度，通过阅读量、转发量、回帖量、活跃度来衡量。阅读量通过网络文本的点击数量来测量，转发量通过网络文本在全互联网范围内不同网站的出现次数来测量，回帖量通过网络文本回复数量来测量，活跃度通过单位时间内对网络文本的回复数量来测量③受众态度指的是浏览者对网络文本所表达的观点的认同度，通过正面态度回帖数量、中性态度回帖数量、负面态度回帖数量来衡量。The audience attention index aims to reflect the influence of Internet public opinion on the audience, and is measured by indicators such as audience situation, audience response, and audience attitude. ①Audience status refers to the number and scope of audiences affected by public opinion. The number of audiences is measured by the number of IPs of Internet text viewers, and the range of audiences is measured by the geographical distribution of IPs of Internet text viewers. ②Audience response refers to the degree of attention of viewers to online texts, which is measured by the amount of reading, reposting, replies, and activity. The amount of reading is measured by the number of clicks on the web text, the amount of forwarding is measured by the number of times the web text appears on different websites across the Internet, the amount of replies is measured by the number of replies to the web text, and the activity is measured by the replies to the web text per unit time Quantitative measurement ③Audience attitude refers to the viewer's recognition of the views expressed in the network text, which is measured by the number of positive, neutral and negative replies.
该指标体系的各级指标的权重通过层次分析法计算得出，每一项指标均可量化计算得出，指标的量化计算方法分为三种：指数计算、频率/密度计算和权重系数确定。The weights of the indicators at all levels of the indicator system are calculated through the AHP, and each indicator can be quantified. The quantitative calculation methods of the indicators are divided into three types: index calculation, frequency/density calculation and weight coefficient determination.
(1)指数计算(1) Index calculation
指标体系中有定量指标和定性指标。定量指标包括阅读量、转发量、回帖量等指标；定性指标包括视听化程度。为具有可比性，将定性指标与定量指标按归一化处理，这里采用指数计算方法，具体采用Sigmoid函数进行计算，其中x表示阅读量、转发量、回帖量等。以受众响应为例，设对于网络文本i，网络文本的点击数量为x1i，网络文本在全互联网范围内不同网站的出现次数为x2i，网络文本回复数量为x3i，单位时间内对网络文本的回复数量为x4i。设阅读量、转发量、回帖量、活跃度的权重是g1、g2、g3、g4，则网络文本对受众响应的影响力P1为：There are quantitative indicators and qualitative indicators in the indicator system. Quantitative indicators include indicators such as reading volume, forwarding volume, and reply volume; qualitative indicators include the degree of audio-visualization. In order to have comparability, the qualitative index and the quantitative index are normalized, and the index calculation method is used here, specifically the Sigmoid function Perform calculations, where x represents the amount of reading, reposting, and replies, etc. Taking audience response as an example, suppose that for web text i, the number of clicks on the web text is x1i , the number of appearances of the web text on different websites in the entire Internet is x2i , and the number of replies to the web text is x3i . The number of replies to the text is x4i . Assuming that the weights of reading volume, forwarding volume, reply volume, and activity are g1, g2, g3, and g4, then the influence P1 of online texts on audience responses is:
P1＝f(x1i)×g1+f(x2i)×g2+f(x3i)×g3+f(x4i)×g4P1＝f(x1i )×g1 +f(x2i )×g2 +f(x3i )×g3 +f(x4i )×g4
(2)频率计算(2) Frequency calculation
活跃度是根据网民对网络文本的回复频率来衡量，以天、星期、月为统计时间单位。Activity is measured according to the frequency of Internet users' replies to texts on the Internet, with days, weeks, and months as statistical time units.
(3)权重系数确定(3) Determination of weight coefficient
根据专家经验利用层次分析法确定各种属性因素的权重系数。其主要特征是把复杂的问题分解为若干个组成因素，将这些因素按从属关系分为层次结构；专家评比时只需对各因素进行两两比较，确定同一层次中诸因素的相对重要性，然后综合专家的判断决定各因素相对重要的顺序。用这种方法来决定各因素的加权系数比在很多因素中凭经验同时定出加权系数更科学一些，因为人们只作两两比较时容易得出比较准确的判断。但在使用这些方法时，为了保证效果，每一层次所包含的因素一般超过10个。进行两两对比时按9分制进行，1代表相当，3是稍好，5是明显地好，7是十分好，9是极好。如介于上述二者之间则用2、4、6或8分表示。根据两两对比打分结果构成评分矩阵，通过求矩阵的最大特征根和特征向量即可计算出各因素相对于上一层目标的重要性或评价权重。如果要求计算各参数对再上一层目标的重要性顺序或影响程度大小，可以将底层的各参数的权重逐一乘上与其有关的上一层因素的权重，然后相加，这样各参数对再上一层的优劣顺序或加权系数就计算出来了。According to expert experience, the weight coefficients of various attribute factors are determined by AHP. Its main feature is to decompose complex problems into several constituent factors, and divide these factors into hierarchical structures according to their affiliation; experts only need to compare each factor in pairs to determine the relative importance of factors in the same level. Then comprehensive expert judgment determines the order of relative importance of each factor. Using this method to determine the weighting coefficients of each factor is more scientific than setting the weighting coefficients based on experience among many factors at the same time, because it is easier for people to make more accurate judgments when they only make pairwise comparisons. However, when using these methods, in order to ensure the effect, each level generally contains more than 10 factors. A 9-point scale is used for pairwise comparisons, with 1 being equivalent, 3 being slightly better, 5 being significantly better, 7 being very good, and 9 being excellent. If it is between the above two, it is expressed with 2, 4, 6 or 8 points. The scoring matrix is constructed according to the scoring results of the pairwise comparison, and the importance or evaluation weight of each factor relative to the target of the previous layer can be calculated by finding the largest eigenvalue and eigenvector of the matrix. If it is required to calculate the order of importance or the degree of influence of each parameter on the target of the next layer, the weight of each parameter at the bottom layer can be multiplied by the weight of the factor of the upper layer related to it one by one, and then added, so that each parameter has an impact on the next layer. The pros and cons or weighting coefficients of the previous layer are calculated.
量化评估结果的计算公式为,The formula for calculating the quantitative evaluation results is,
其中，Ai代表一级指标，舆情强度和受众关注度的分值，ωi代表各自的权重。Among them, Ai represents the first-level index, the score of public opinion strength and audience attention, and ωi represents their respective weights.
每一个一级指标则是由其下属的二级指标决定的，计算公式为其中，是第i个一级指标的第j项，其权重为ωj。类似地，每一个二级指标由其下属的三级指标决定。Each first-level indicator is determined by its subordinate second-level indicators, and the calculation formula is in, is the j-th item of the i-th first-level index, and its weight is ωj . Similarly, each second-level indicator is determined by its subordinate third-level indicators.
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510243751.0ACN104820629B (en) | 2015-05-14 | 2015-05-14 | A kind of intelligent public sentiment accident emergent treatment system and method |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510243751.0ACN104820629B (en) | 2015-05-14 | 2015-05-14 | A kind of intelligent public sentiment accident emergent treatment system and method |
| Publication Number | Publication Date |
|---|---|
| CN104820629Atrue CN104820629A (en) | 2015-08-05 |
| CN104820629B CN104820629B (en) | 2018-01-30 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510243751.0AActiveCN104820629B (en) | 2015-05-14 | 2015-05-14 | A kind of intelligent public sentiment accident emergent treatment system and method |
| Country | Link |
|---|---|
| CN (1) | CN104820629B (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105956740A (en)* | 2016-04-19 | 2016-09-21 | 北京深度时代科技有限公司 | Semantic risk calculating method based on text logical characteristic |
| CN106202561A (en)* | 2016-07-29 | 2016-12-07 | 北京联创众升科技有限公司 | Digitized contingency management case library construction methods based on the big data of text and device |
| CN106294619A (en)* | 2016-08-01 | 2017-01-04 | 上海交通大学 | Public sentiment intelligent supervision method |
| CN107193796A (en)* | 2016-03-14 | 2017-09-22 | 北大方正集团有限公司 | A kind of public sentiment event detecting method and device |
| CN107239452A (en)* | 2016-03-28 | 2017-10-10 | 腾讯科技（深圳）有限公司 | A kind of method and device of Developing Tactics |
| CN107274324A (en)* | 2017-06-06 | 2017-10-20 | 张黎明 | A kind of method that accident risk assessment is carried out based on cloud service |
| CN107491438A (en)* | 2017-08-25 | 2017-12-19 | 前海梧桐（深圳）数据有限公司 | Business decision elements recognition method and its system based on natural language |
| CN107590196A (en)* | 2017-08-15 | 2018-01-16 | 中国农业大学 | Earthquake emergency information screening and evaluating system and system in a kind of social networks |
| CN107622354A (en)* | 2017-09-29 | 2018-01-23 | 中国科学技术大学 | A Method of Emergency Response Capability Evaluation Based on Interval Binary Semantics |
| CN107741929A (en)* | 2017-10-18 | 2018-02-27 | 网智天元科技集团股份有限公司 | The analysis of public opinion method and device |
| CN108108902A (en)* | 2017-12-26 | 2018-06-01 | 阿里巴巴集团控股有限公司 | A kind of risk case alarm method and device |
| CN108255832A (en)* | 2016-12-28 | 2018-07-06 | 航天信息股份有限公司 | public sentiment processing system and method |
| CN109615266A (en)* | 2018-12-26 | 2019-04-12 | 贵州电网有限责任公司 | The text analyzing decision-making technique of power grid exception information based on data mining |
| CN110046220A (en)* | 2018-12-13 | 2019-07-23 | 阿里巴巴集团控股有限公司 | Public feelings information processing method, device, equipment and computer readable storage medium |
| CN110096406A (en)* | 2018-01-31 | 2019-08-06 | 阿里巴巴集团控股有限公司 | A kind of event of failure discovery method and server |
| CN110609969A (en)* | 2019-08-08 | 2019-12-24 | 阿里巴巴集团控股有限公司 | Information processing method and device |
| CN110852090A (en)* | 2019-11-07 | 2020-02-28 | 中科天玑数据科技股份有限公司 | Public opinion crawling mechanism characteristic vocabulary extension system and method |
| CN110868383A (en)* | 2018-12-24 | 2020-03-06 | 北京安天网络安全技术有限公司 | Website risk assessment method and device, electronic equipment and storage medium |
| CN111223026A (en)* | 2020-01-03 | 2020-06-02 | 武汉理工大学 | Intelligent management method of waste crisis transformation |
| CN111428146A (en)* | 2020-03-24 | 2020-07-17 | 上海智臻智能网络科技股份有限公司 | Network information processing method and system, equipment and storage medium |
| CN111898385A (en)* | 2020-07-17 | 2020-11-06 | 中国农业大学 | An earthquake disaster assessment method and system |
| CN112069381A (en)* | 2020-09-27 | 2020-12-11 | 中国科学院深圳先进技术研究院 | A monitoring management method and system based on natural language processing technology |
| CN112905745A (en)* | 2021-03-05 | 2021-06-04 | 广州虎牙科技有限公司 | Information processing method, information processing device, electronic equipment and computer readable storage medium |
| WO2021136453A1 (en)* | 2019-12-31 | 2021-07-08 | 清华大学 | Method and apparatus for obtaining emergency plan auxiliary information, and device |
| CN113433994A (en)* | 2021-07-20 | 2021-09-24 | 青岛一云数联科技有限公司 | System and method for sensing and monitoring energy event |
| CN113468320A (en)* | 2021-07-22 | 2021-10-01 | 中国地震台网中心 | Method and system for quickly visualizing earthquake emergency information |
| CN113626722A (en)* | 2020-05-08 | 2021-11-09 | 国家广播电视总局广播电视科学研究院 | Public opinion guiding method, device, equipment and computer readable storage medium |
| CN113961668A (en)* | 2021-10-22 | 2022-01-21 | 曾水求 | A smart campus management system based on big data |
| CN114186118A (en)* | 2021-12-08 | 2022-03-15 | 深圳市网联安瑞网络科技有限公司 | A network public opinion topic information processing system, method, storage medium and terminal |
| CN114417021A (en)* | 2022-01-24 | 2022-04-29 | 中国电子科技集团公司第五十四研究所 | An accurate distribution method of semantic information based on the fusion of time, space and semantic constraints |
| CN114444514A (en)* | 2022-02-08 | 2022-05-06 | 北京百度网讯科技有限公司 | Semantic matching model training method, semantic matching method and related device |
| CN114528396A (en)* | 2021-12-29 | 2022-05-24 | 北京辰安科技股份有限公司 | Method and device for monitoring emergency, electronic equipment and storage medium |
| CN117076752A (en)* | 2023-03-21 | 2023-11-17 | 南京时刻智能科技有限公司 | Data information acquisition and analysis method based on big data |
| CN117891997A (en)* | 2023-12-29 | 2024-04-16 | 浙江翎鹿数智科技有限公司 | Public opinion management and cross-department joint treatment method and system |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101436194A (en)* | 2008-11-04 | 2009-05-20 | 中国电子科技集团公司第五十四研究所 | Text multiple-accuracy representing method based on data excavating technology |
| CN101819573A (en)* | 2009-09-15 | 2010-09-01 | 电子科技大学 | Self-adaptive network public opinion identification method |
| CN102509164A (en)* | 2011-11-24 | 2012-06-20 | 广州市地下铁道总公司 | Automatic generation method for digital emergency plan |
| CN103150335A (en)* | 2013-01-25 | 2013-06-12 | 河南理工大学 | Co-clustering-based coal mine public sentiment monitoring system |
| CN103544255A (en)* | 2013-10-15 | 2014-01-29 | 常州大学 | Text semantic relativity based network public opinion information analysis method |
| CN104573016A (en)* | 2015-01-12 | 2015-04-29 | 武汉泰迪智慧科技有限公司 | System and method for analyzing vertical public opinions based on industry |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101436194A (en)* | 2008-11-04 | 2009-05-20 | 中国电子科技集团公司第五十四研究所 | Text multiple-accuracy representing method based on data excavating technology |
| CN101819573A (en)* | 2009-09-15 | 2010-09-01 | 电子科技大学 | Self-adaptive network public opinion identification method |
| CN102509164A (en)* | 2011-11-24 | 2012-06-20 | 广州市地下铁道总公司 | Automatic generation method for digital emergency plan |
| CN103150335A (en)* | 2013-01-25 | 2013-06-12 | 河南理工大学 | Co-clustering-based coal mine public sentiment monitoring system |
| CN103544255A (en)* | 2013-10-15 | 2014-01-29 | 常州大学 | Text semantic relativity based network public opinion information analysis method |
| CN104573016A (en)* | 2015-01-12 | 2015-04-29 | 武汉泰迪智慧科技有限公司 | System and method for analyzing vertical public opinions based on industry |
| Title |
|---|
| 马力: "基于Web2.0的政府网络舆情信息管理模型研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》* |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107193796A (en)* | 2016-03-14 | 2017-09-22 | 北大方正集团有限公司 | A kind of public sentiment event detecting method and device |
| CN107193796B (en)* | 2016-03-14 | 2021-12-24 | 北大方正集团有限公司 | Public opinion event detection method and device |
| CN107239452A (en)* | 2016-03-28 | 2017-10-10 | 腾讯科技（深圳）有限公司 | A kind of method and device of Developing Tactics |
| CN105956740A (en)* | 2016-04-19 | 2016-09-21 | 北京深度时代科技有限公司 | Semantic risk calculating method based on text logical characteristic |
| CN105956740B (en)* | 2016-04-19 | 2019-12-31 | 北京深度时代科技有限公司 | Semantic risk calculation method based on text logical features |
| CN106202561B (en)* | 2016-07-29 | 2019-10-01 | 北京联创众升科技有限公司 | Digitlization contingency management case base construction method and device based on text big data |
| CN106202561A (en)* | 2016-07-29 | 2016-12-07 | 北京联创众升科技有限公司 | Digitized contingency management case library construction methods based on the big data of text and device |
| CN106294619A (en)* | 2016-08-01 | 2017-01-04 | 上海交通大学 | Public sentiment intelligent supervision method |
| CN108255832A (en)* | 2016-12-28 | 2018-07-06 | 航天信息股份有限公司 | public sentiment processing system and method |
| CN107274324A (en)* | 2017-06-06 | 2017-10-20 | 张黎明 | A kind of method that accident risk assessment is carried out based on cloud service |
| CN107590196A (en)* | 2017-08-15 | 2018-01-16 | 中国农业大学 | Earthquake emergency information screening and evaluating system and system in a kind of social networks |
| CN107491438A (en)* | 2017-08-25 | 2017-12-19 | 前海梧桐（深圳）数据有限公司 | Business decision elements recognition method and its system based on natural language |
| CN107622354B (en)* | 2017-09-29 | 2020-06-26 | 中国科学技术大学 | An emergency response capability assessment method based on interval binary semantics |
| CN107622354A (en)* | 2017-09-29 | 2018-01-23 | 中国科学技术大学 | A Method of Emergency Response Capability Evaluation Based on Interval Binary Semantics |
| CN107741929A (en)* | 2017-10-18 | 2018-02-27 | 网智天元科技集团股份有限公司 | The analysis of public opinion method and device |
| CN108108902B (en)* | 2017-12-26 | 2021-06-29 | 创新先进技术有限公司 | Risk event warning method and device |
| CN108108902A (en)* | 2017-12-26 | 2018-06-01 | 阿里巴巴集团控股有限公司 | A kind of risk case alarm method and device |
| CN110096406A (en)* | 2018-01-31 | 2019-08-06 | 阿里巴巴集团控股有限公司 | A kind of event of failure discovery method and server |
| CN110046220A (en)* | 2018-12-13 | 2019-07-23 | 阿里巴巴集团控股有限公司 | Public feelings information processing method, device, equipment and computer readable storage medium |
| CN110868383A (en)* | 2018-12-24 | 2020-03-06 | 北京安天网络安全技术有限公司 | Website risk assessment method and device, electronic equipment and storage medium |
| CN109615266B (en)* | 2018-12-26 | 2022-11-04 | 贵州电网有限责任公司 | Text analysis decision method for power grid abnormal information based on data mining |
| CN109615266A (en)* | 2018-12-26 | 2019-04-12 | 贵州电网有限责任公司 | The text analyzing decision-making technique of power grid exception information based on data mining |
| CN110609969A (en)* | 2019-08-08 | 2019-12-24 | 阿里巴巴集团控股有限公司 | Information processing method and device |
| CN110852090A (en)* | 2019-11-07 | 2020-02-28 | 中科天玑数据科技股份有限公司 | Public opinion crawling mechanism characteristic vocabulary extension system and method |
| CN110852090B (en)* | 2019-11-07 | 2024-03-19 | 中科天玑数据科技股份有限公司 | Mechanism characteristic vocabulary expansion system and method for public opinion crawling |
| WO2021136453A1 (en)* | 2019-12-31 | 2021-07-08 | 清华大学 | Method and apparatus for obtaining emergency plan auxiliary information, and device |
| CN111223026A (en)* | 2020-01-03 | 2020-06-02 | 武汉理工大学 | Intelligent management method of waste crisis transformation |
| CN111223026B (en)* | 2020-01-03 | 2024-03-01 | 武汉理工大学 | Intelligent management method for garbage crisis transformation |
| CN111428146A (en)* | 2020-03-24 | 2020-07-17 | 上海智臻智能网络科技股份有限公司 | Network information processing method and system, equipment and storage medium |
| CN113626722A (en)* | 2020-05-08 | 2021-11-09 | 国家广播电视总局广播电视科学研究院 | Public opinion guiding method, device, equipment and computer readable storage medium |
| CN111898385A (en)* | 2020-07-17 | 2020-11-06 | 中国农业大学 | An earthquake disaster assessment method and system |
| CN111898385B (en)* | 2020-07-17 | 2023-08-04 | 中国农业大学 | Earthquake disaster assessment method and system |
| CN112069381A (en)* | 2020-09-27 | 2020-12-11 | 中国科学院深圳先进技术研究院 | A monitoring management method and system based on natural language processing technology |
| CN112905745A (en)* | 2021-03-05 | 2021-06-04 | 广州虎牙科技有限公司 | Information processing method, information processing device, electronic equipment and computer readable storage medium |
| CN113433994A (en)* | 2021-07-20 | 2021-09-24 | 青岛一云数联科技有限公司 | System and method for sensing and monitoring energy event |
| CN113468320A (en)* | 2021-07-22 | 2021-10-01 | 中国地震台网中心 | Method and system for quickly visualizing earthquake emergency information |
| CN113961668A (en)* | 2021-10-22 | 2022-01-21 | 曾水求 | A smart campus management system based on big data |
| CN114186118A (en)* | 2021-12-08 | 2022-03-15 | 深圳市网联安瑞网络科技有限公司 | A network public opinion topic information processing system, method, storage medium and terminal |
| CN114528396A (en)* | 2021-12-29 | 2022-05-24 | 北京辰安科技股份有限公司 | Method and device for monitoring emergency, electronic equipment and storage medium |
| CN114417021B (en)* | 2022-01-24 | 2023-08-25 | 中国电子科技集团公司第五十四研究所 | Semantic information accurate distribution method based on time, space and sense multi-constraint fusion |
| CN114417021A (en)* | 2022-01-24 | 2022-04-29 | 中国电子科技集团公司第五十四研究所 | An accurate distribution method of semantic information based on the fusion of time, space and semantic constraints |
| CN114444514B (en)* | 2022-02-08 | 2023-01-24 | 北京百度网讯科技有限公司 | Semantic matching model training method, semantic matching method and related device |
| CN114444514A (en)* | 2022-02-08 | 2022-05-06 | 北京百度网讯科技有限公司 | Semantic matching model training method, semantic matching method and related device |
| CN117076752A (en)* | 2023-03-21 | 2023-11-17 | 南京时刻智能科技有限公司 | Data information acquisition and analysis method based on big data |
| CN117891997A (en)* | 2023-12-29 | 2024-04-16 | 浙江翎鹿数智科技有限公司 | Public opinion management and cross-department joint treatment method and system |
| CN117891997B (en)* | 2023-12-29 | 2024-09-13 | 浙江翎鹿数智科技有限公司 | Public opinion management and cross-department joint treatment method and system |
| Publication number | Publication date |
|---|---|
| CN104820629B (en) | 2018-01-30 |
| Publication | Publication Date | Title |
|---|---|---|
| CN104820629B (en) | A kind of intelligent public sentiment accident emergent treatment system and method | |
| CN110008311B (en) | A product information security risk monitoring method based on semantic analysis | |
| CN113378565B (en) | Event analysis method, device, device and storage medium for multi-source data fusion | |
| Chen et al. | Crime prediction using Twitter sentiment and weather | |
| CN103699626B (en) | Method and system for analysing individual emotion tendency of microblog user | |
| CN110516067A (en) | Public opinion monitoring method, system and storage medium based on topic detection | |
| CN103854063B (en) | A kind of prediction of event occurrence risk method for early warning based on internet opening imformation | |
| CN103399891A (en) | Method, device and system for automatic recommendation of network content | |
| CN103854064B (en) | Event occurrence risk prediction and early warning method targeted to specific zone | |
| CN105045857A (en) | Social network rumor recognition method and system | |
| CN111310476A (en) | Public opinion monitoring method and system using aspect-based emotion analysis method | |
| CN104346425A (en) | Method and system of hierarchical internet public sentiment indication system | |
| CN103853700B (en) | A kind of event method for early warning found based on region and object information | |
| CN104899335A (en) | Method for performing sentiment classification on network public sentiment of information | |
| CN110717654A (en) | Product quality evaluation method and system based on user comments | |
| Alsaedi et al. | A combined classification-clustering framework for identifying disruptive events | |
| Duong et al. | Neurosymbolic ai for mining public opinions about wildfires | |
| CN114969560B (en) | Crisis information pushing method and system based on event situation and user cognition | |
| Subramani et al. | Text mining and real-time analytics of twitter data: A case study of australian hay fever prediction | |
| CN109885760B (en) | Information traceability method and system based on user interest | |
| CN117828076A (en) | Public opinion grading early warning method and system based on propagation link | |
| CN114169524B (en) | A Media Relationship Management Expert System | |
| Deshpande et al. | BI and sentiment analysis | |
| CN116450924A (en) | The analysis module and method of the ecological environment public opinion monitoring system of the intelligent environmental protection platform | |
| CN114547167A (en) | Automobile public opinion sentiment analysis method |
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| EXSB | Decision made by sipo to initiate substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |