WO2022151524A1 - 一种利用社交软件大数据找寻潜在客户的方法 - Google Patents

一种利用社交软件大数据找寻潜在客户的方法 Download PDF

Info

Publication number
WO2022151524A1
WO2022151524A1 PCT/CN2021/073532 CN2021073532W WO2022151524A1 WO 2022151524 A1 WO2022151524 A1 WO 2022151524A1 CN 2021073532 W CN2021073532 W CN 2021073532W WO 2022151524 A1 WO2022151524 A1 WO 2022151524A1
Authority
WO
WIPO (PCT)
Prior art keywords
model
customers
screening
data
screening model
Prior art date
Application number
PCT/CN2021/073532
Other languages
English (en)
French (fr)
Inventor
幸格曼吉特
高登
严诚
Original Assignee
中山世达模型制造有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 中山世达模型制造有限公司 filed Critical 中山世达模型制造有限公司
Publication of WO2022151524A1 publication Critical patent/WO2022151524A1/zh

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/953—Querying, e.g. by the use of web search engines
    • G06F16/9535—Search customisation based on user profiles and personalisation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism

Definitions

  • the invention relates to a method for finding potential customers by utilizing the big data of social software.
  • the present invention provides a method for establishing social software big data to find potential customers.
  • a method for finding potential customers by using big data of social software which is characterized by comprising the following steps.
  • S1 create screening model 1 and screening model 2 based on historical customer information as data.
  • the filter conditions include educational background, education field, current employer, tenure, position, whether to use any 3D or CAD software, the filter model one automatically searches the Internet according to the filter conditions. conditions for new customers.
  • the information of the new customers searched by the screening model 1 is input into the screening model 2, and the industry and company size are added as the screening conditions of the screening model 2 to screen again.
  • the creation of the first screening model and the second screening model includes the following steps.
  • S1.1 data collection, select a certain number of customers from historical customers and collect the information of these customers as data.
  • Customers include customers who have placed orders and customers who sent inquiries to our company but did not place orders in the end.
  • the information includes education Background, field of education, current company, years of employment, title held, use of any 3D or CAD software.
  • Model Evaluation 1 evaluates the random forest by checking the confusion matrix, recall results and specificity, and selects the one with the highest accuracy as the screening model 1.
  • model test 1 randomly select a certain number of customers in social networking sites as a test, and judge the effect of screening model 1 by the success rate.
  • the screening model 2 is a random forest model identical to the screening model 1.
  • the output data of the screening model 1 is used as the input data of the screening model 2, and the industry and company size are added as features.
  • model test 2 improve the accuracy of screening model 2 through hyperparameter adjustment and complete the establishment of the model.
  • the data analysis also includes analyzing all variables and creating graphs to analyze the data, and building a machine learning model based on the data to find correlations between various features.
  • the chart includes a list of positions, a list of the size of the company currently employed, and a list of educational fields.
  • the screening model 1 can automatically search for potential customers that meet the conditions according to the set screening conditions, and after further screening by the screening model 2, the customized information is sent to the customer, and customers with a relatively high success rate can be screened out.
  • our method works more efficiently, converts more and saves time, thereby avoiding wasting time on unsuitable contacts and investing time and energy in more suitable prospects.
  • Figure 1 is an analysis graph of job types.
  • Figure 2 is an analysis diagram of educational types.
  • Figure 3 is an analysis graph of company size.
  • Figure 4 is a flow chart of the present invention.
  • the present invention discloses a method for finding potential customers using big data of social software, including the following steps.
  • S1 create a screening model 1 and a screening model 2 based on the historical customer information as data.
  • the filter conditions include educational background, education field, current employer, tenure, position, whether to use any 3D or CAD software, the filter model one automatically searches the Internet according to the filter conditions.
  • the new customers of the conditions can be one of the above undergraduate, mechanical major, Midea Group, working for more than three years, management, using 3D or CAD software, so that we can preliminarily screen out the required customers according to the above conditions.
  • the information of the new customers searched by the screening model 1 is input into the screening model 2, and the industry and company size are added as the screening conditions of the screening model 2 to screen again.
  • the industry and company size are added as the screening conditions of the screening model 2 to screen again.
  • the creation of the first screening model and the second screening model includes the following steps.
  • Historical customers include customers who have placed orders and customers who sent inquiries to our company but did not place orders in the end.
  • the information includes educational background, Education field, current company, tenure, position held, use of any 3D or CAD software, we mixed the information of these customers in a database, and then we collected the information they disclosed on social software.
  • the analysis process includes analyzing all the variables and creating charts to analyze the data, using the data to build a machine learning model, these data all contain different characteristics, so through these charts we get more data that helps us determine the distribution and variables of the target variable
  • the relationship between tells us how many of the contacts of the customer who placed the order are engineers or procurement (as shown in Figure 1, the abscissa indicates the position, and the ordinate indicates the number of people corresponding to the position); how many of the customer's contacts have educational qualifications It is a master's or doctoral degree (as shown in Figure 2, the abscissa represents the educational level, and the ordinate represents the number of people corresponding to the degree); how many customers are there (as shown in Figure 3, the abscissa represents the quantity range, and the ordinate represents the number of people) The revenue corresponding to this quantity range), etc.
  • the target variable can distinguish customers who have placed an order from those who have not.
  • the distribution of this variable is approximately 49% to 51%. 49% of the customers in the database have placed an order and 51% of the customers have not placed an order.
  • model creation create multiple random forest models
  • model creation requires training set first, followed by test set, training random forest model through training set and testing random forest model through test set, we can use 60:40, 70 : 30 or 80: 20 ratio, preferably 70: 30, these data are enough for model testing, according to the total amount of data, randomly divided into two groups according to the ratio, the data set is randomly divided, and the two sets contain various variables . data according to After the 70-30 ratio split, we created the random forest model.
  • a random forest model distributes data in a random fashion and builds many decision trees, then democratically takes the average of all decision trees. Random forest requires less computing power, it is easy to deploy, and it does exactly what we need.
  • Model Evaluation 1 evaluates the random forest by checking the confusion matrix, recall results and specificity, and selects the one with the highest accuracy as the screening model 1.
  • model test 1 randomly select a certain number of customers from social networking sites as a test, and judge the effect of screening model 1 by the success rate.
  • the screening model 2 is a random forest model that is the same as the screening model 1.
  • the output data of the screening model 1 is used as the input data of the screening model 2 and the industry and company size are added as features.
  • the screening model The overall accuracy of one is about 82%. But the recall is (1, 0---62, 98). This means that the model was 98% accurate at screening unsuitable people, but only 62% accurate at selecting matched customers.
  • establish screening model two which will screen out more people who are less likely to place an order with us.
  • screening model two we need more specific features to separate potential customers from non-customers, so, We decided to use the two characteristics of industry and company size. As a manufacturing company, certain industries and company sizes are very suitable for us. Therefore, the purpose of the second model is to filter out the output of the first model again. The rest of the unsuitable personnel.
  • Model Evaluation 2 detects the accuracy of the screening model 2, and evaluates the accuracy of the screening model 2 by checking the confusion matrix, recall results and specificity.
  • the accuracy of the screening model 2 reaches 85%, which is significantly higher than the first model. improvement.
  • model test 2 improve the accuracy of screening model 2 through hyperparameter adjustment and complete the establishment of the model, and finally increase the accuracy of the model to 89%.
  • the artificial intelligence customer source filter built according to the principle of this method is a sales efficiency software, which can cooperate with LinkedIn, Zhuo Information (American B2B marketing intelligent service company), Xingwang (German business social networking site), etc.
  • the use of any type of market development tool can significantly reduce the spam generated by the company's sales and marketing departments through more accurate customer source screening.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Databases & Information Systems (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Software Systems (AREA)
  • Data Mining & Analysis (AREA)
  • Business, Economics & Management (AREA)
  • Primary Health Care (AREA)
  • Medical Informatics (AREA)
  • Tourism & Hospitality (AREA)
  • Strategic Management (AREA)
  • Marketing (AREA)
  • Artificial Intelligence (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Human Resources & Organizations (AREA)
  • Evolutionary Computation (AREA)
  • General Business, Economics & Management (AREA)
  • General Health & Medical Sciences (AREA)
  • Computing Systems (AREA)
  • Economics (AREA)
  • Mathematical Physics (AREA)
  • Health & Medical Sciences (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

本发明公开了一种利用社交软件大数据找寻潜在客户的方法，本方法通过筛选模型一能够自动按照设置的筛选条件搜寻符合条件的潜在客户，再通过筛选模型二的进一步筛选后发送定制信息给该客户可以筛选出成功率比较高的客户，相比之前人工筛选和联系，我们的方法工作效率高、转化率高、节省时间，从而避免浪费时间在不适合的联系人，将时间和精力投入更适合的潜在客户。

Description

一种利用社交软件大数据找寻潜在客户的方法 技术领域
本发明涉及一种利用社交软件大数据找寻潜在客户的方法。
 背景技术
开发新客户是业务发展最重要的因素之一。但是，寻找新客户是一个非常困难的过程，它需要大量时间和金钱的投资，而且仍然不能保证新客户的顺利开发。
市场上有很多付费软件可以提供潜在客户的信息，这些信息基于几个条件（地理位置，行业等），之后你可以尝试通过电话、电子邮件等联系这些潜在客户。电话销售或发送电子邮件是开发新客户的传统方法，但通常效果不大，转化率很低。因为没有针对真正有需要的潜在客户进行发送，因而大部分的状况是，对方根本不回复或不接你的电话，因而目前我们迫切需要一种利用社交软件大数据快捷准确筛选出有需要的潜在客户的方法。
 技术解决方案
为了克服现有技术的不足，本发明提供一种建立社交软件大数据找寻潜在客户的方法。
本发明解决其技术问题所采用的技术方案是。
一种利用社交软件大数据找寻潜在客户的方法，其特征在于：包括以下步骤。
 S1, 依据历史客户的信息作为数据创建筛选模型一和筛选模型二。
S2, 对筛选模型一设置筛选条件，筛选条件包括教育背景、教育领域、目前就职公司、任职年限、出任职务、是否使用任何 3D 或CAD 软件，筛选模型一依据筛选条件自动在网络上搜寻符合筛选条件的新客户。
S3, 经过筛选模型一搜寻的新客户的信息输入筛选模型二同时附加行业及公司规模做为筛选模型二的筛选条件再次筛选。
S4, 符合条件则发送定制信息联系该新客户。
所述筛选模型一和筛选模型二的创建包括以下步骤。
   S1.1，数据收集，从历史客户中挑选一定数量的客户并收集这些客户的信息作为数据，客户包括下过订单的客户和给我司发送询价但最终没有下订单的客户，信息包括教育背景、教育领域、目前就职公司、任职年限、出任职务、是否使用任何 3D 或CAD 软件。
S1.2，数据清洗，丢弃无法复制数据的行。
S1.3，数据分析，寻找各种特征之间的相关性。
S1.4，特征工程， 将部分特征分类为变量，并将变量转换为数值。
S1.5，模型创建，创建多个随机森林模型。
S1.6，模型评估一，通过检查混淆矩阵、召回结果和特异性来评估随机森林，选出准确性最高的作为筛选模型一。
S1.7，模型测试一，在社交网站中随机选取一定数量的客户作为测试，通过成功率判断筛选模型一的效果。
S1.8，创建筛选模型二，筛选模型二为一个与筛选模型一相同的随机森林模型，将筛选模型一的输出数据作为筛选模型二的输入数据并且附加行业及公司规模做为特征。
S1.9，模型评估二，检测筛选模型二的精度。
S1.10，模型测试二，通过超参数调整提高筛选模型二的精度并完成模型的建立。
所述数据分析还包括分析所有变量和创建图表解析数据，并且依据数据建立机器学习模型找出各种特征之间的相关性。
所述图表包括职务列表、目前就职公司规模列表、学历领域列表。
 有益效果
本发明的有益效果是：筛选模型一能够自动按照设置的筛选条件搜寻符合条件的潜在客户，再通过筛选模型二的进一步筛选后发送定制信息给该客户可以筛选出成功率比较高的客户，相比之前人工筛选和联系，我们的方法工作效率高、转化率高、节省时间，从而避免浪费时间在不适合的联系人，将时间和精力投入更适合的潜在客户。
 附图说明
下面结合附图和实施例对本发明进一步说明。
图1是职位类型的分析图。
图2是学历类型的分析图。
图3是公司规模的分析图。
图4是本发明的流程图。
具体实施方式。
参照图1至图4，本发明公开了一种利用社交软件大数据找寻潜在客户的方法，包括以下步骤。
S1, 依据历史客户的信息作为数据创建筛选模型一和筛选模型二。
S2, 对筛选模型一设置筛选条件，筛选条件包括教育背景、教育领域、目前就职公司、任职年限、出任职务、是否使用任何 3D 或CAD 软件，筛选模型一依据筛选条件自动在网络上搜寻符合筛选条件的新客户，如筛选条件可以是本科以上、机械专业、美的集团、工作三年以上、管理层、使用3D 或CAD 软件其中的一种，这样根据上述条件我们能够初步筛选出所需要的客户。
S3, 经过筛选模型一搜寻的新客户的信息输入筛选模型二同时附加行业及公司规模做为筛选模型二的筛选条件再次筛选，筛选条件为家电行业、公司营收超一千万或公司人数超一百人等。
S4, 符合条件则发送定制信息联系该新客户。
所述筛选模型一和筛选模型二的创建包括以下步骤。
  S1.1，数据收集，挑选2000 名历史客户并收集这些历史客户的信息作为数据，历史客户包括下过订单的客户和给我司发送询价但最终没有下订单的客户，信息包括教育背景、教育领域、目前就职公司、任职年限、出任职务、是否使用任何 3D 或CAD 软件，我们将这些客户的信息混合在数据库里，然后我们收集了他们在社交软件上公开的信息。
S1.2，数据清洗，因为收集到的数据是混乱的，因此，我们在适当的使用了插补方法清理数据，并丢弃了几行无法复制数据的行。
S1.3，数据分析，在此过程中，我们对整个数据进行分析，并寻找各种特征之间的相关性，确定哪些特征是重要变量，确定各种数据、目标变量的分布。
分析过程包括分析所有变量和创建图表解析数据，采用数据构建了机器学习模型，这些数据背后都包含不同的特征，因此，通过这些图表我们获得了更多数据，帮助我们确定目标变量的分布和变量之间的关系。例如，它告诉我们下订单的客户的联系人中有多少人是工程师或采购（如附图1所示，横坐标表示职位，纵坐标表示该职位对应的人数）；多少客户的联系人的学历是硕士或博士（如附图2所示，横坐标表示学历，纵坐标表示该学历对应的人数）；多少客户的规模有多大（如附图3所示，横坐标表示数量范围，纵坐标表示该数量范围对应的营收）等。然后我们对数据进行了分析，为了找出这些特征，我们检查了这些特征与目标变量的相关性，我们使用内置的导入变量功能，构建了一个基本模型并绘制出一个相关的图表，这个图表可以告诉我们哪些变量最重要，我将这些变量按升序进行排列。
目标变量可以区分下过订单与未下订单的客户。此变量的分布约为 49% 至51%。数据库中 49% 的客户下过订单， 51% 的人客户未下过订单。
S1.4，特征工程，部分特征都被分类为变量，这些变量格式各不相同。要使用这些变量，使用了独热编码技术将它们转换为数值。
S1.5，模型创建，创建多个随机森林模型，模型创建首先需要训练集，之后是测试集，通过训练集训练随机森林模型和通过测试集测试随机森林模型，我们可以用60：40，70：30或80：20的比率，优选了70：30，这些数据足够进行模型测试，根据数据的总量，按照比率随机分成两组，数据集被随机划分，两个集合中包含了各种变量。将数据按照 70-30 比率划分后，我们创建了随机森林模型。
随机森林模型以随机的方式分配数据，并建造出很多决策树，随后以民主方式采取所有决策树的平均值。随机森林所需的计算能力较少，它易于部署，完全可以满足我们的要求。
S1.6，模型评估一，通过检查混淆矩阵、召回结果和特异性来评估随机森林，选出准确性最高的作为筛选模型一。
S1.7，模型测试一，在社交网站中随机选取一定数量的客户作为测试，通过成功率判断筛选模型一的效果，我们使用领英网站中约500人的原始信息进行测试。筛选模型一在运行这些数据后，为我们筛选了 140 名潜在客户，然后我们在领英中发送建立联系的邀请给这些客户，以前人工筛选的邀请的接受率在20%左右，但现在达到了80%左右，因此，我们看到模型即时提升了转换率，帮助我们从原始数据中找出更适合的潜在客户。
S1.8，创建筛选模型二，筛选模型二为一个与筛选模型一相同的随机森林模型，将筛选模型一的输出数据作为筛选模型二的输入数据并且附加行业及公司规模做为特征，筛选模型一的整体精度约为 82%。但召回是（1，0---62，98）。这意味着模型在筛除不适合人员时准确率达到 98%，但在选择匹配客户时的准确率只有 62%。在此基础上建立筛选模型二，筛选模型二将筛除更多不太可能向我们下订单的人员，在筛选模型二中，我们需要更具体的特征，将潜在客户与非客户分开，因此，我们决定使用行业及公司规模这两个特征，作为一家加工制造型企业，某些行业和公司规模非常适合我们，因此，第二个模型的目的就是从第一个模型的输出中，再次过滤掉其余不适合的人员。
S1.9，模型评估二，检测筛选模型二的精度，通过检查混淆矩阵、召回结果和特异性来评估筛选模型二的精度，筛选模型二的精度达到了 85%，比第一个模型有显著的提高。
S1.10，模型测试二，通过超参数调整提高筛选模型二的精度并完成模型的建立，最终将模型的准确度提升到了89%。
综上所述，依据本方法原理建成的人工智能客源筛选器是一款销售效率软件，可以和领英、瞩目信息(美国B2B营销智能化服务公司）、行网（德国商务社交网站）等任何类型的市场开发工具匹配使用，通过更为精准的客源筛选，显著减少公司销售和市场部门产生的垃圾邮件。
以上对本发明实施例所提供的一种利用社交软件大数据找寻潜在客户的方法进行了详细介绍，本文中应用了具体个例对本发明的原理及实施方式进行了阐述，以上实施例的说明只是用于帮助理解本发明的方法及其核心思想；同时，对于本领域的一般技术人员，依据本发明的思想，在具体实施方式及应用范围上均会有改变之处，综上所述，本说明书内容不应理解为对本发明的限制。

Claims (5)

  1. 一种利用社交软件大数据找寻潜在客户的方法，其特征在于：包括以下步骤：
       S1, 依据历史客户的信息作为数据创建筛选模型一和筛选模型二，
    S2, 对筛选模型一设置筛选条件，筛选条件包括教育背景、教育领域、目前就职公司、任职年限、出任职务、是否使用任何 3D 或CAD 软件，筛选模型一依据筛选条件自动在网络上搜寻符合筛选条件的新客户，
    S3, 经过筛选模型一搜寻的新客户的信息输入筛选模型二同时附加行业及公司规模做为筛选模型二的筛选条件再次筛选，
    S4, 符合条件则发送定制信息联系该新客户。
  2. 根据权利要求1所述的一种利用社交软件大数据找寻潜在客户的方法，其特征在于：所述筛选模型一和筛选模型二的创建包括以下步骤：
       S1.1，数据收集，从历史客户中挑选一定数量的客户并收集这些客户的信息作为数据，客户包括下过订单的客户和给我司发送询价但最终没有下订单的客户，信息包括教育背景、教育领域、目前就职公司、任职年限、出任职务、是否使用任何 3D 或CAD 软件，
    S1.2，数据清洗，丢弃无法复制数据的行，
    S1.3，数据分析，寻找各种特征之间的相关性，
    S1.4，特征工程， 将部分特征分类为变量，并将变量转换为数值,
    S1.5，模型创建，创建多个随机森林模型，
    S1.6，模型评估一，通过检查混淆矩阵、召回结果和特异性来评估随机森林，选出准确性最高的作为筛选模型一，
    S1.7，模型测试一，在社交网站中随机选取一定数量的客户作为测试，通过成功率判断筛选模型一的效果，
    S1.8，创建筛选模型二，筛选模型二为一个与筛选模型一相同的随机森林模型，将筛选模型一的输出数据作为筛选模型二的输入数据并且附加行业及公司规模做为特征，
    S1.9，模型评估二，检测筛选模型二的精度，
    S1.10，模型测试二，通过超参数调整提高筛选模型二的精度并完成模型的建立。
  3. 根据权利要求2所述的一种利用社交软件大数据找寻潜在客户的方法，其特征在于：所述数据分析还包括分析所有变量和创建图表解析数据，并且依据数据建立机器学习模型找出各种特征之间的相关性。
  4. 根据权利要求1所述的一种利用社交软件大数据找寻潜在客户的方法，其特征在于：所述图表包括职务列表、目前就职公司规模列表、学历领域列表。
  5. 根据权利要求1所述的一种利用社交软件大数据找寻潜在客户的方法，其特征在于：S1.5中的模型创建还包括创建训练集和测试集，通过训练集训练随机森林模型和通过测试集测试随机森林模型。
PCT/CN2021/073532 2021-01-14 2021-01-25 一种利用社交软件大数据找寻潜在客户的方法 WO2022151524A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202110046743.2 2021-01-14
CN202110046743.2A CN112667911A (zh) 2021-01-14 2021-01-14 一种利用社交软件大数据找寻潜在客户的方法

Publications (1)

Publication Number Publication Date
WO2022151524A1 true WO2022151524A1 (zh) 2022-07-21

Family

ID=75415145

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/073532 WO2022151524A1 (zh) 2021-01-14 2021-01-25 一种利用社交软件大数据找寻潜在客户的方法

Country Status (2)

Country Link
CN (1) CN112667911A (zh)
WO (1) WO2022151524A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117910804A (zh) * 2024-01-31 2024-04-19 广东寰讯信息股份有限公司 基于大数据和多模态智能算法的风险防范感知方法及系统

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113177151A (zh) * 2021-05-28 2021-07-27 中山世达模型制造有限公司 一种潜在客户筛选方法

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106779827A (zh) * 2016-12-02 2017-05-31 上海晶樵网络信息技术有限公司 一种互联网用户行为采集及分析检测的大数据方法
CN110009416A (zh) * 2019-04-02 2019-07-12 安徽筋斗云机器人科技股份有限公司 一种基于大数据清洗和ai精准营销的系统
US10846352B1 (en) * 2013-12-23 2020-11-24 Massachusetts Mutual Life Insurance Company System and method for identifying potential clients from aggregate sources

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105488697A (zh) * 2015-12-09 2016-04-13 焦点科技股份有限公司 一种基于客户行为特征的潜在客户挖掘方法
CN106228389A (zh) * 2016-07-14 2016-12-14 武汉斗鱼网络科技有限公司 基于随机森林算法的网络潜力用户挖掘方法及系统
US20180225685A1 (en) * 2017-02-07 2018-08-09 Linkedin Corporation Identifying impending user-competitor relationships on an online social networking system
CN108256052B (zh) * 2018-01-15 2023-07-11 成都达拓智通科技有限公司 基于tri-training的汽车行业潜在客户识别方法
CN110222272B (zh) * 2019-04-18 2022-10-14 广东工业大学 一种潜在客户挖掘与推荐方法

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10846352B1 (en) * 2013-12-23 2020-11-24 Massachusetts Mutual Life Insurance Company System and method for identifying potential clients from aggregate sources
CN106779827A (zh) * 2016-12-02 2017-05-31 上海晶樵网络信息技术有限公司 一种互联网用户行为采集及分析检测的大数据方法
CN110009416A (zh) * 2019-04-02 2019-07-12 安徽筋斗云机器人科技股份有限公司 一种基于大数据清洗和ai精准营销的系统

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117910804A (zh) * 2024-01-31 2024-04-19 广东寰讯信息股份有限公司 基于大数据和多模态智能算法的风险防范感知方法及系统

Also Published As

Publication number Publication date
CN112667911A (zh) 2021-04-16

Similar Documents

Publication Publication Date Title
CN107220353B (zh) 一种智能客服机器人满意度自动评价方法及系统
CN105335400A (zh) 针对用户的提问意图获取答案信息的方法及装置
WO2022151524A1 (zh) 一种利用社交软件大数据找寻潜在客户的方法
CN109710766A (zh) 一种工单数据的投诉倾向分析预警方法及装置
CN113435627A (zh) 基于工单轨迹信息的电力客户投诉预测方法及装置
CN113723635A (zh) 一种面向客服系统的智能派单方法
Manresa et al. Investigating the impact of new technologies and organizational practices on operational performance: evidence from Spanish manufacturing companies
CN115456745A (zh) 小微企业画像构建方法及装置
CN113379432B (zh) 一种基于机器学习的销售系统客户匹配方法
He et al. A study on prediction of customer churn in fixed communication network based on data mining
Lu et al. The selection of construction material suppliers in supplier relationship management (SRM)
CN113240209A (zh) 一种基于图神经网络的城市产业集群发展路径预测方法
CN116151840B (zh) 一种基于大数据的用户服务数据智能管理系统及方法
CN117035283A (zh) 用于投标人员配置的人力资源信息获取方法及系统
CN110727801A (zh) 基于本体的模糊评价搜索引擎广告优化方法
CN115994765A (zh) 一种基于云服务的crm管理开发系统
Dragan Fuzzy-DEA model for measuring the efficiency of transport quality
Sopadang et al. The role of lean sustainable logistics in business operations
Yuan et al. Critical risks identification of Public Private Partnerships in China and the analysis on questionnaire survey
CN113962581A (zh) 一种基于大数据的产业链工厂产能评估方法及系统
Tabaroki et al. Identification and prioritisation of the critical success factors for research project-based organisations using fuzzy analytic hierarchy process
Xu et al. Big data analytics toward intelligent mobile service provisions of customer relationship management in e-commerce
Hermansyah et al. Implementation of Participatory Rural Appraisal (PRA) in Empowering Gaplek SMEs Using Partial Least Square (PLS) Analysis.
CN117172729B (zh) 一种基于大数据的劳务分包人员管理系统
CN115130007B (zh) 一种基于用户场景定位的品牌推广方法及系统

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21918723

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21918723

Country of ref document: EP

Kind code of ref document: A1

122 Ep: pct application non-entry in european phase

Ref document number: 21918723

Country of ref document: EP

Kind code of ref document: A1