JP2005149489A - Program, and business operation support system and method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To effectively support the business operation of a user by increasing kinds of items of business operation confirmable by the user. <P>SOLUTION: One embodiment of this application is a program 2b for realizing: a function 14 for reading business operation data 71-7n created by associating at least either of business item data showing business items predefined in a computer 1 and freely inputted free text data, time data and business operation identification data with one another; a function 15 for providing the free text data to a means 3 for extracting a predetermined item preregistered in dictionary data based on the text data when the free text data are included in the business operation data 71-7n; a function 16 for receiving the extracted item extracted from the free text data by the means 3; and a function 17 for creating business operation transition data 261 expressing the business items and the extracted items in chronological order by using the business operation data having the same business operation identification data. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a program for supporting sales activities, a sales activity support system, and a method.

  The sales activity support device stores sales activity information related to past sales activities, and is used by the user to refer to the past sales activity information in the current sales activity.

  For example, in Patent Document 1 (Japanese Patent Laid-Open No. 2001-250022), activity plan data closed by a case is accumulated in a success pattern database, and success pattern data that is closely related to the activity plan data for each ongoing case is extracted. A sales activity support device is disclosed.

  Patent Document 1 describes that business activities are divided into business items (categories) that are subdivided in advance, such as “search, grasp needs, business negotiation, proposal”. And the sales activity support apparatus of patent document 1 displays the progress for every business item which was subdivided beforehand.

Patent Document 2 discloses a method of inputting objective information about a customer, categorizing the customer, and appropriately distributing sales force (see, for example, paragraphs 39 to 42).
Japanese Patent Laid-Open No. 2001-250022 Japanese Patent No. 3234744

  In the sales activity support device of Patent Document 1, the sales activity is subdivided into sales items in advance, and the user confirms the progress of each predefined sales item.

  Therefore, in this patent document 1, the user can confirm the progress of the business items defined in advance among the business activities.

  However, in Patent Document 1, it is difficult for the user to confirm the progress of other items that are not defined in advance in the business activities.

  Further, Patent Document 1 has a description that success pattern data that is closely related to ongoing activity plan data is extracted.

  However, Patent Document 1 does not disclose in what case the relationship between the ongoing activity plan data and the success pattern data is deeply related.

  Further, the business activity support device of Patent Document 1 outputs an alarm when the difference between the activity plan data in progress and the success pattern data is equal to or greater than a predetermined value.

  However, it may be difficult for a user who has received an alarm to determine what kind of activity should be performed after the alarm is output.

  On the other hand, in Patent Document 2, there is a description that a future visit plan is created based on a target value for ongoing activity data (see, for example, paragraphs 43 and 48). Patent Document 2 normally has no particular problem, but according to the study of the present inventor, it does not disclose a method for extracting an existing business activity that can be used as a reference for future business activities. In addition, Patent Document 2 has no particular problem as described above, but according to the study of the present inventor, there is no disclosure of a technique that can classify activity data in progress in the same way as completed sales activities.

  The present invention has been made in view of the above circumstances, and a program for more effectively supporting a user's sales activities by increasing the types of items of sales activities that can be confirmed by the user, and The purpose is to provide a sales activity support system and method.

  Another object of the present invention is to provide a program that can extract an existing business activity that can be used as a reference for a future business activity with respect to the ongoing business activity.

  Furthermore, another object of the present invention is to provide a program that can classify sales activities in progress in the same way as completed sales activities.

  Specific means taken for realizing the present invention will be described below.

  According to the first embodiment of the present invention, the computer includes the business item data indicating any of the business items defined in advance as being implemented in the business activity and the free text data freely input regarding the business activity. A function for reading out a plurality of sales activity data associated with at least one of the data, time data, and sales activity identification data for identifying the sales activity, and free text data is included in any of the sales data. A function that provides free text data to a text mining means that extracts predetermined items registered in dictionary data in advance based on the text data, and is extracted from the free text data by the text mining means. Function that accepts extracted items, sales activity data with the same sales activity identification data Used, a program for realizing a function of creating a sales activity transition data representing the operating items and extracted item by time sequence based on time data.

  In the second embodiment, at least one of sales item data indicating any one of the sales items defined in advance as being executed in the sales activity and free text data freely input regarding the sales activity is stored in the computer. A function that reads multiple sales activity data that associates data, time data, and sales activity identification data that identifies a sales activity from the recording means, and free text data is included in one of the multiple sales activity data A function for providing free text data to text mining means for extracting predetermined items registered in dictionary data in advance based on text data, extracted items extracted from free text data by text mining means Using the sales activity data with the same function and sales activity identification data Function to create multiple sales activity transition data that represents sales items and extracted items in time series based on time data, other sales activity transition data similar to the predetermined sales activity transition data for multiple sales activity transition data And a similar output function for outputting the extracted business activity transition data.

  In the third embodiment, at least one of sales item data indicating any of the sales items defined in advance as being executed in the sales activity and free text data freely input regarding the sales activity is stored in the computer. A function that reads multiple sales activity data that associates data, time data, and sales activity identification data that identifies a sales activity from the recording means, and free text data is included in one of the multiple sales activity data A function for providing free text data to text mining means for extracting predetermined items registered in dictionary data in advance based on text data, extracted items extracted from free text data by text mining means Using the sales activity data with the same function and sales activity identification data A function that creates multiple sales activity transition data representing sales items and extracted items in time series based on time data, and shows the transition relationship between sales items and extracted items based on multiple sales activity transition data This is a program for realizing the function of creating a model.

  In the fifth embodiment, at least one of business item data indicating any of the business items defined in advance as being implemented in a business activity and free text data freely input with respect to the business activity is stored in the computer. A function of reading out a plurality of sales activity data in association with the data, time data, and sales activity identification data for identifying the sales activity from the recording means, and the free text data is included in any of the plurality of sales activity data A function for providing the free text data to a text mining means for extracting a predetermined item previously registered in the dictionary data based on the text data, if included, the free text by the text mining means Function for accepting extracted items extracted from data, the business activity identification data Using the same sales activity data, based on the sales item and the extracted item, a function of extracting a plurality of feature amount data from a plurality of completed sales activity data, the plurality of feature amount data, Based on the sales item and the extracted item, the function for classifying into the category represented by the sales item data included in each completed sales activity data and the sales activity data having the same sales activity identification data are used. A function of extracting feature quantity data from the sales activity data, a function of classifying feature quantity data related to the incomplete sales activity data into categories based on a classification result of each of the completed sales activity data, and the incomplete sales activity This is a program for realizing a classification result output function for outputting a data classification result.

  According to the present invention, it is possible to show the progress of sales activities by combining predefined sales items and items corresponding to expressions included in free text data. For this reason, it is possible to increase the contents of sales activities that can be confirmed by the user, and to support the user's sales activities more effectively.

  In addition, by extracting other sales activity transition data similar to the predetermined sales activity transition data and outputting the extracted sales activity transition data, it will be a reference for future sales activities regarding the ongoing sales activities. Existing sales activities can be extracted.

  Furthermore, feature quantity data is extracted from incomplete sales activity data, and feature quantity data related to incomplete sales activity data is classified into categories based on the classification results of each completed sales activity data. By outputting the result of classifying the activity data, it is possible to classify the ongoing sales activities in the same way as completed sales activities.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. In the following embodiments, the same portions are denoted by the same reference numerals, description thereof is omitted, and only different portions will be described in detail.

(First embodiment)
In the present embodiment, a sales activity support system that creates sales activity transition data indicating the progress of sales activities by combining predefined sales items and extracted items extracted from free text data will be described.

  In this embodiment, a unit obtained by subdividing sales activities is defined as a sales item.

  FIG. 1 is a block diagram showing an example of a sales activity support system according to the present embodiment.

  The computer 1 reads and executes the programs 2a and 2b recorded on the recording medium 2.

  Then, the computer 1 operates according to the program 2a and realizes the function as the text mining system 3. Further, the computer 1 operates in accordance with the program 2b and realizes a function as the sales activity support system 4.

  The text mining system 3 accesses the recording unit 5. The recording unit 5 may be a recording device such as an internal memory or a hard disk provided in the computer 1, a recording device external to the computer 1, or a recording device accessed from the computer 1 via a network. A set of a plurality of recording devices may be used.

  The recording unit 5 records various data such as analysis target data and dictionary data 6. In the present embodiment, it is assumed that sales activity data 71 to 7n for reporting the contents of sales activities is recorded in the recording unit 5 as an example of analysis target data.

  FIG. 2 is a diagram showing an example of sales activity data 71 to 7n according to the present embodiment.

  Each sales activity data 71 to 7n includes time data 8 and sales activity identification data 9, respectively.

  Each business activity data 71 to 7n includes at least one of business item data 10 and free text data 11.

  The time data 8 is data indicating time related to sales activities. For example, the time data 8 may be date / time data created for the sales activity data 71 to 7n. Further, the time data 8 may be date and time data when the business item indicated by the business item data 10 is completed.

  The sales activity identification data 9 is data for identifying the sales activity. In this embodiment, it is assumed that the sales activity identification data 9 includes sales staff data 9a, department data 9b, customer name data 9c, and business case name data 9d.

  The sales item data 10 is data indicating a sales item selected by the user 12 among a plurality of sales items for each of the sales activity data 71 to 7n.

  As business items, it is assumed that important items to be confirmed in business activities are defined in advance, such as “quotation presentation”, “suggestion”, “order receipt”, and “losing order”.

  The free text data 11 is data indicating a report of free sales activities by the user 12.

  FIG. 3 is a diagram illustrating an example of the dictionary data 6. In the dictionary data 6, an expression to be extracted from text data by text mining, a sales activity item corresponding to the expression to be extracted, and an item ID corresponding to the item are registered in an associated state. ing.

  The expression registered in the dictionary data 6 is, for example, an expression effective for grasping the progress state of the sales activity.

  For example, in the dictionary data 6, expressions that affect the subjectivity of the user 12 “Demonstration is unpopular” “Demonstration response is slow” “No questions were asked” “Demonstration was well received” “Positive questions were raised” The expressions “Another company withdraws”, “Rivals withdraw bids”, “Discounts”, “DC”, “Discount”, etc., are registered.

  In this embodiment, an expression is a combination of one or more characters having a certain meaning. For example, expressions include words, phrases, clauses, sentences, and the like.

  In addition, the dictionary data 6 includes the high-level items that are unified with the registered expressions, such as “Demonstration unpopular”, “Demonstration favorable reception”, “Extract from other companies”, “Discount”, etc., in a state associated with each expression. Yes.

  For example, the items “Demonstration is unpopular” and the expressions “Demonstration is unpopular”, “Demonstration is slow to respond”, “No questions” are related.

  Also, the item “Demonstration is well received” and the expressions “Demonstration is well received” and “A positive question has been issued” are related.

  Also, the item “withdrawal from other companies” and the expressions “withdrawal from other companies” and “competitor withdraws bids” are related.

  The item “discount” is associated with the expressions “discount”, “DC”, and “discount”.

  Further, in the dictionary data 6, items “Demonstration unpopular”, “Demonstration favorable reception”, “Extract from other companies”, “Discount”, etc., and item IDs “G001” to “G004” are registered in an associated state. Yes.

  The text mining system 3 realizes functions as a data input unit 3a, an analysis unit 3b, and a data output unit 3c.

  The sales activity support system 4 includes a registration unit 13, a reading unit 14, a provision unit 15, an extraction item reception unit 16, a transition creation unit 17, an inconsistency determination unit 18, an output control unit 19, a designation reception unit 20, and a selection unit 21. Realize the function. The output control unit 19 includes a mismatch output unit 19a, a transition output unit 19b, and a detailed output unit 19c.

  First, the configuration of the text mining system 3 will be described.

  In the present embodiment, the data input unit 3a of the text mining system 3 receives the free text data 11 included in the sales activity data 71 to 7n from the providing unit 15 of the sales activity support system 4, and the free text data 11 Is provided to the analysis unit 3b.

  When the analysis unit 3b receives the free text data 11 from the data input unit 3a, the analysis unit 3b reads the dictionary data 6 recorded in the recording unit 5.

  In addition, the analysis unit 3 b determines whether the expression registered in the dictionary data 6 is registered in the free text data 11.

  When the free text data 11 includes an expression registered in the dictionary data 6, the analysis unit 3 b includes items and item IDs corresponding to the expressions existing in the free text data 11 and the dictionary data 6. To extract.

  Then, the analysis unit 3b provides the output unit 3c with extracted data 22 including the extracted item (hereinafter referred to as an extracted item) and an item ID. The extracted data 22 may include an extracted item, an item ID, and contents indicating from which free text data the extracted item is extracted.

  The data output unit 3c records the extracted data 22 including the extracted item and the item ID in the recording unit 5.

  Next, the configuration of the sales activity support system 4 will be described.

  The registration unit 13 of the sales activity support system 4 displays a sales activity data registration screen on the output device 23 such as a display.

  FIG. 4 is a diagram illustrating an example of a sales activity data registration screen.

  The sales activity data registration screen 24 includes a time data input unit 24a for inputting time data 8, a salesperson data input unit 24b for inputting salesperson data 9a, and a department for inputting department data 9b. A data input unit 24c, a customer name data input unit 24d for inputting customer name data 9c, and a business case name data input unit 24e for inputting business case name data 9d are provided.

  The sales activity data registration screen 24 includes a business item selection unit 24f used for selecting a business item defined in advance, and a free text data input unit 24g for inputting the free text data 11.

  The registration unit 13 accepts the content input by the user 12 on the sales activity data registration screen 24 by operating the input device 25 such as a keyboard or a pointing device.

  And the registration part 13 records the sales activity data 71-7n in the recording part 5 based on the input content.

  When receiving the sales activity support command from the input device 25 based on the operation of the user 12, the reading unit 14 reads the business activity data 71 to 7 n recorded in the recording unit 5. Then, the reading unit 14 provides the business activity data 71 to 7 n to the providing unit 15.

  Further, the reading unit 14 provides the business activity data 71 to 7 n to the transition creation unit 17.

  When receiving the business activity data 71 to 7n from the reading unit 14, the providing unit 15 provides the free text data 11 included in the business activity data 71 to 7n to the data input unit 3a of the text mining system 3. .

  The extracted item receiving unit 16 receives the extracted data 22 recorded in the recording unit 5 as a result of providing the free text data 11 included in the business activity data 71 to 7n to the data input unit 3a of the text mining system 3.

  Then, the extraction item reception unit 16 provides the extraction data 22 to the transition creation unit 17.

  The transition creation unit 17 receives the business activity data 71 to 7n from the reading unit 14 and receives the extraction data 22 from the extraction item reception unit 16.

  In addition, the transition creation unit 17 sets the combination of the time data 8 and the business item indicated by the business item data 10 and the time data 8 for each business activity identification data 9 based on the business activity data 71 to 7n and the extracted data 22. And sales activity transition data 261 to 26m in which combinations of the extracted items included in the extracted data 22 are arranged in time series.

  Then, the transition creation unit 17 records the business activity transition data 261 to 26 m in the recording unit 5.

  FIG. 5 is a diagram illustrating an example of a creation state of the sales activity transition data 261 to 26 m by the transition creation unit 17.

  In the example of FIG. 5, the sales activity data 71 includes time data “January 20”, sales activity identification data “B1”, and sales item data “suggestion”.

  The sales activity data 7a includes time data “January 31” and sales activity identification data “B2”. In addition, from the sales activity data 7a, an item “start activity” corresponding to the expression “start activity” in the free text data is extracted.

  The sales activity data 7b includes time data “February 18” and sales activity identification data “B1”. Further, an item “discount” corresponding to the expression “discount” in the free text data is extracted from the sales activity data 7b.

  The sales activity data 7c includes time data “March 3”, sales activity identification data “B2”, and sales item data “presentation”.

  The sales activity data 7d includes time data “April 4”, sales activity identification data “B1”, and sales item data “order received”.

  The sales activity data 7n includes time data “May 5”, sales activity identification data “B2”, and sales item data “lost”.

  The transition creation unit 17 uses a combination of time data and business items “January 20: Proposal”, “April 4: Order received” from the business activity data 71, 7b, 7d including the same business activity identification data “B1”. ”And a combination of time data and extracted items“ February 18: Discount ”, and sales activity transition data 261 in which the created combinations are arranged in time series is created.

  Similarly, the transition creation unit 17 uses the combination of time data and business items “March 3: Estimate” “May” from the business activity data 7a, 7c, 7n including the same business activity identification data “B2”. “5th: Unsuccessful” and a combination of time data and extracted item “January 31: Activity start” are created, and sales activity transition data 26m is created in which the created combinations are arranged in time series.

  The inconsistency determination unit 18 determines whether or not the contents match between the sales item and the extracted item based on the business activity transition data 261 to 26m.

  When the inconsistency determining unit 18 determines that there is an inconsistency, the inconsistency determining unit 18 provides the inconsistent output unit 19a of the output control unit 19 with data indicating an inconsistent portion in the sales activity transition data in which the inconsistency occurs. To do.

  FIG. 6 is a diagram illustrating an example of a consistency determination method performed by the inconsistency determination unit 18.

  For example, at the time of registration of certain sales activity data 7e, the user 12 selects the sales item “Demonstration favorable” from among the predefined sales items “Demonstration favorable reception”, “Demonstration response normal”, “Demonstration unpopular”, and the like. At the same time, it is assumed that the free text data “There was no question” was input.

  The transition creation unit 17 then combines a combination of the time data “October 11” and the sales item “Demonstration favorable” “October 11: Demonstration favorable” based on some sales activity data including the sales activity data 7e. And “October 11: Demonstration unpopular” combination of extracted items “Demonstration unpopular” corresponding to the expression “No questions” input to the time data “October 11” and the free text data. Assume that data 26a is created.

  For example, the inconsistency determination unit 18 refers to the table 27 in which the contents of the business items defined in advance are associated with items that do not match the contents of the business items.

  For example, in the table 27, it is assumed that a predefined business item “Demonstration is popular” and an item “Demonstration is not popular” are associated with each other. The table 27 is recorded in the recording unit 5, for example.

  In this case, the inconsistency determination unit 18 determines that the extracted item “Demonstration Unpopular” does not match the business item “Demonstration Unpopular”.

  If the inconsistency determining unit 18 determines that there is an inconsistency, the inconsistency output unit 19a of the output control unit 19 sends the data 28 indicating the inconsistent portion in the sales activity transition data in which the inconsistency occurs. provide.

  The inconsistency determination unit 18 supplies the inconsistency output unit 19a with data indicating a total result such as the number of occurrences of inconsistency for each salesperson or customer based on the salesperson data 9a or the customer name data 9c. May be provided.

  In the example of FIG. 6, the inconsistent portion is represented by the time data “October 11”.

  Based on the data 28, the mismatch output unit 19a of the output control unit 19 performs processing for highlighting and displaying the mismatch portion on the output device 23. The inconsistency output unit 19a receives data indicating the inconsistency aggregation result for each sales person or department, and performs processing for causing the output device 23 to display the aggregation result.

  The transition output unit 19b reads the business activity transition data 261 to 26m recorded in the recording unit 5.

  And the transition output part 19b is based on the sales activity transition data 261-26m, The process for displaying on the output device 23 the graph which set the 1st axis as a time axis and has arrange | positioned the sales item or the extraction item to the 2nd axis. I do.

  When the user 12 operates the input device 25 and inputs a detailed display command for the output result of the sales activity transition data 261 to 26m, the designation receiving unit 20 receives the detailed display command from the user 12 and provides it to the selecting unit 21 To do.

  When receiving the detailed display command from the designation receiving unit 20, the selection unit 21 selects business activity data corresponding to the part designated to be displayed in detail for the user 12 and reads it from the recording unit 5. The sales activity data corresponding to the designated portion can be selected by using, for example, time data 8, sales item data 10, and free text data 11.

  Then, the selection unit 21 provides the selected sales activity data to the detailed output unit 19c.

  The detailed output unit 19c performs processing for causing the output device 23 to display the selected business activity data or a part thereof.

  Below, an example of the output state of the sales activity transition data 261-26m by the output control part 19 is demonstrated.

  FIG. 7 is a diagram illustrating an example of an output result by the transition output unit 19b.

  In FIG. 7, sales activity transition data is displayed with the horizontal axis as the time axis. In addition, on the vertical axis, an extraction item corresponding to a pre-defined business item indicated by the business item data 10 or an expression extracted from the free text data 11 is arranged.

  In FIG. 7, for example, it is assumed that the extracted item “Demonstration favorable reception” is an extracted item corresponding to an expression included in the free text data 11.

  In FIG. 7, the customer name is displayed according to the customer name data 9 c, “order received” or “missed” is displayed as the final result, and the activity period represented by the sales activity transition data is displayed.

  FIG. 8 is a diagram illustrating a first example of an output result by the mismatch output unit 19a.

  In FIG. 8, the vertical axis and the horizontal axis are the same as those in FIG. In FIG. 8, there is a discrepancy between the sales item “Demonstration Unpopular” that is defined in advance and selected by the user 12 and the extracted item “Demonstration Positive” corresponding to the expression extracted from the free text data 11. . For this reason, the portion where this inconsistency occurs is highlighted.

  FIG. 9 is a diagram illustrating a second example of an output result by the mismatch output unit 19a.

  In FIG. 9, the graph of FIG. 8 is displayed in a table format, and the portion corresponding to the date “4/20” where the inconsistency occurs is highlighted.

  FIG. 10 is a diagram illustrating an example of an output result by the detailed output unit 19c.

  FIG. 10 illustrates a case where the date “5/30” portion of the graph of FIG. 7 is designated by the user 12. The selection unit 21 is based on the displayed sales activity transition data, the extracted item corresponding to the expression extracted from the sales item data 10 or the free text data 11, and the time data 8 corresponding to the date designated by the user 12. , Sales activity data corresponding to the designation of the user 12 is selected from the sales activity data 71 to 7n of the recording unit 5. The detailed output unit 19c displays the selected sales activity data.

  11 and 12 are flowcharts showing an example of processing by the business activity support system 4 and the text mining system 3 according to the present embodiment.

  In step S <b> 1, the registration unit 13 of the sales activity support system 4 displays a sales activity data registration screen 24 on the output device 23.

  In step S <b> 2, the registration unit 13 accepts the content input by the user 12 on the sales activity data registration screen 24 by operating the input device 25, and records the sales activity data 71 to 7 n in the recording unit 5.

  In step S <b> 3, the reading unit 14 receives a business activity support command from the input device 25 based on the operation of the user 12.

  In step S <b> 4, the reading unit 14 reads the business activity data 71 to 7 n recorded in the recording unit 5 and provides the business activity data 71 to 7 n to the providing unit 15 and the transition creating unit 17.

  In step S <b> 5, the providing unit 15 provides the free text data 11 included in the business activity data 71 to 7 n to the data input unit 3 a of the text mining system 3.

  In step S <b> 6, the data input unit 3 a of the text mining system 3 receives the free text data 11 included in the sales activity data 71 to 7 n from the providing unit 15 of the sales activity support system 4 and analyzes the free text data 11. Part 3b.

  In step S7, the analysis unit 3b receives the free text data 11 from the data input unit 3a and reads the dictionary data 6 recorded in the recording unit 5.

  In step S8, the analysis unit 3b extracts the expression registered in the dictionary data 6 from the free text data 11, and provides an extraction item corresponding to the extracted expression to the data output unit 3c.

  In step S9, the data output unit 3c records the extracted data 22 including the extracted items in the recording unit 5.

  In step S <b> 10, the extracted item receiving unit 16 of the sales activity support system 4 receives the extracted data 22 recorded in the recording unit 5 and provides the extracted data 22 to the transition creating unit 17.

  In step S <b> 11, the transition creation unit 17 creates a combination of the time data 8 and the business item indicated by the business item data 10 for each business activity identification data 9 based on the business activity data 71 to 7n and the extracted data 22. Sales activity transition data 261 to 26 m in which combinations of time data 8 and extracted items are arranged in time series is created and recorded in the recording unit 5.

  In step S12, the inconsistency determination unit 18 determines whether or not the contents match between the sales item and the extracted item based on the business activity transition data 261 to 26m.

  If the inconsistency determining unit 18 determines that there is inconsistency, in step S13a, the inconsistent output unit 19a performs a process for displaying sales activity transition data highlighting the inconsistent portion on the output device 23. Do.

  If the inconsistency determining unit 18 determines that there is no inconsistency, the transition output unit 19b performs processing for displaying the business activity transition data on the output device 23 in step S13b.

  In step S <b> 14, the designation receiving unit 20 determines whether a detailed display command for the output result of the business activity transition data is input from the user 12 who operates the input device 25.

  If no detail display command is input, the process ends.

  When the detailed display command is input, in step S15, the selection unit 21 corresponds to a portion designated to be displayed in detail to the user 12 from the sales activity data 71 to 7n recorded in the recording unit 5. Select sales activity data.

  In step S <b> 16, the detailed output unit 19 c performs processing for causing the output device 23 to display the selected business activity data or a part thereof. After execution of step S16, the processing from step S14 onward is repeated.

  In the present embodiment described above, the user 12 is provided with the progress of the sales activities that complement the contents of the free text data 11 described freely, rather than the progress of the sales activities only for the predefined sales items. .

  Therefore, the user 12 can investigate the progress of the business activity more concretely and easily.

  Moreover, in this Embodiment, the consistency of the content described in the business item defined beforehand and the free text data 11 can be confirmed.

  Therefore, it is possible to grasp the accuracy of the input of sales activity data. In addition, by obtaining the result of consistency calculation for each salesperson or department, it is possible to instruct the salesperson or department in which many inconsistencies have occurred to correctly input sales activity data.

  In the present embodiment, the business activity transition data based on the business items defined in advance and the extracted items corresponding to the expressions included in the free text data 11 corresponds to the portion designated by the user 12. Sales activity data is displayed.

  Therefore, after grasping the transition of the sales activity, the user 12 can examine the details of the sales activity in detail.

  Further, in the present embodiment, the free text data 11 is checked for the presence of expressions registered in the dictionary data 6, and there is an overlapping expression between the free text data 11 and the dictionary data 6. The item and item ID corresponding to this expression are extracted. The subsequent processing is executed using unified items or item IDs instead of various expression units.

  Thereby, even if it is a case where an extraction item is added to the business item defined beforehand, subsequent processing can be simplified and the user's 12 understanding degree and grasping degree can be improved about the progress of sales activity.

(Second Embodiment)
In the present embodiment, a modification of the first embodiment will be described.

  In the present embodiment, a sales activity support system for obtaining sales activity transition data related to a completed sales activity that is similar to sales activity transition data related to an incomplete sales activity will be described.

  FIG. 13 is a block diagram showing an example of a sales activity support system according to the present embodiment.

  The computer 1 reads and executes the programs 2a and 2c recorded on the recording medium 29.

  The computer 1 operates according to the program 2c and realizes a function as the sales activity support system 30.

  The sales activity support system 30 is similar to the sales activity support system 4 according to the first embodiment described above. The registration unit 13, the reading unit 14, the providing unit 15, the extracted item receiving unit 16, the transition creation unit 17, the inconsistency Functions as the determination unit 18, the output control unit 19, the designation receiving unit 20, and the selection unit 21 are realized.

  In addition, the sales activity support system 30 includes a key designation receiving unit 31, a similarity determining unit 32, and a similar output control unit 33.

  In the present embodiment, the user 12 operates the input device 25 and inputs a designation of an incomplete business activity in order to investigate a completed business activity that serves as a reference for the incomplete business activity.

  The key designation accepting unit 31 accepts designation of sales activity transition data related to unfinished business activities input by the user 12 by operating the input device 25.

  Then, the key designation receiving unit 31 provides the similarity determination unit 32 with designation of incomplete business activity transition data.

  The similarity determination unit 32 reads the business activity transition data 261 to 26 m recorded in the recording unit 5 when receiving the designation of incomplete business activity transition data from the key designation receiving unit 31.

  The similarity determination unit 32 selects completed sales activity transition data similar to incomplete sales activity transition data, and creates similarity determination data 34 indicating the selection result.

  Then, the similarity determination unit 32 records the similarity determination data 34 in the recording unit 5.

  FIG. 14 is a diagram illustrating an example of a method for determining the degree of similarity between incomplete sales activity transition data and completed sales activity transition data.

  For the unfinished sales activity transition data 26b, the feature amounts F1 to Fi are obtained based on the sales items that have ended at the current time and the number, type, time interval, and the like of the extracted items extracted at the current time.

  Further, regarding the completed sales activity transition data 26c, the number of sales items in the portion 35 corresponding to the incomplete sales activity transition data 26b of the completed sales activity transition data 26c, and the number, type, and time of the extracted items are extracted. The feature amounts G1 to Gi are obtained based on the interval and the like.

  Then, the smaller the difference between the feature amounts F1 to Fi determined from the incomplete sales activity transition data and the feature amounts G1 to Gi determined from the completed business activity transition data, the higher the similarity is determined. .

  Note that the feature amounts F1 to Fi and the feature amounts G1 to Gi may be weighted according to their contents.

  The similar output control unit 33 reads the similarity determination data 34 recorded in the recording unit 5, and based on the similarity determination data 34, the output device 23 resembles the incomplete sales activity transition data designated by the user 12. Perform processing to display completed sales activity transition data.

  FIG. 15 is a diagram illustrating an example of display of the similarity determination data 34.

  The similar output control unit 33 displays the details of a predetermined number of completed sales activity transition data having a high degree of similarity to the incomplete sales activity transition data, as a result of indicating whether the customer name, the order received, or the lost order. Displays a button that accepts the specification.

  The similar output control unit 33 displays the completed business activity transition data corresponding to the pressed button when the user 12 operates the input device 25 and presses one of the buttons for detailed display. I do.

  FIG. 16 is a flowchart showing an example of similarity determination processing by the business activity support system 30 according to the present embodiment.

  In step T <b> 1, the key specification receiving unit 31 of the sales activity support system 30 receives the specification of incomplete sales activity transition data from the user 12 and provides the similarity determination unit 32 with the specification of incomplete sales activity transition data.

  In step T <b> 2, when the similarity determination unit 32 receives designation of incomplete business activity transition data from the key designation reception unit 31, the similarity determination unit 32 reads the business activity transition data 261 to 26 m recorded in the recording unit 5.

  In step T3, the similarity determination unit 32 selects completed sales activity transition data similar to incomplete sales activity transition data, creates similarity determination data 34 indicating the selection result, and stores the similarity determination data 34 in the recording unit 5. To record.

  In step T <b> 4, the similar output control unit 33 reads out the similarity determination data 34 recorded in the recording unit 5, and based on the similarity determination data 34, the output device 23 sends an incomplete sales activity transition designated by the user 12. Processing for displaying completed sales activity transition data similar to the data is performed.

  In step T <b> 5, the similar output control unit 33 determines whether a detailed display is instructed by the user 12.

  If no detail display command is input, the process ends.

  When the detail display command is input, in step T6, the similar output control unit 33 performs processing for displaying the completed business activity transition data designated by the user 12 to be displayed in detail. After execution of step T6, the processing after step T5 is repeated.

  In the present embodiment described above, completed sales activity transition data similar to incomplete sales activity transition data is displayed.

  Therefore, the user 12 can determine what kind of activity will be performed thereafter with reference to the completed sales activity for the current sales activity. The user 12 can act to bring the current sales activity into the order by referring to the completed sales activity whose result is the order.

  Further, in the present embodiment, as in the first embodiment, not only the business items defined in advance, but also progress on items corresponding to freely described expressions can be confirmed. In particular, the sales activities of the user 12 can be supported.

  In this embodiment, the feature amount is extracted and the completed sales activity transition data similar to the incomplete sales activity transition data is selected. For example, incomplete sales activity transition data that is time-series data is used. Then, the vector operation may be performed on the completed business activity transition data, and the completed business activity transition data determined to be close to the result of the vector operation may be selected. In addition, it is possible to select completed sales activity transition data similar to incomplete sales activity transition data by applying various similar pattern extraction techniques.

  In the present embodiment, data similar to incomplete sales activity transition data is selected from all completed sales activity transition data. However, the sales activity has been completed and the result is “order received”. Data similar to incomplete sales activity transition data may be selected from the transition data. Thereby, the user 12 can refer to how he / she should perform afterwards in order to obtain a favorable result.

(Third embodiment)
In the present embodiment, a modification of the second embodiment will be described.

  In the present embodiment, a sales activity support system that predicts subsequent progress of incomplete sales activity transition data using time series prediction technology will be described.

  FIG. 17 is a block diagram showing an example of a sales activity support system according to the present embodiment.

  The computer 1 reads and executes the programs 2a and 2d recorded on the recording medium 36.

  The computer 1 operates according to the program 2d and realizes a function as the sales activity support system 37.

  The sales activity support system 37 is similar to the sales activity support system 4 according to the first embodiment described above. The registration unit 13, the reading unit 14, the providing unit 15, the extracted item reception unit 16, the transition creation unit 17, the inconsistency Functions as the determination unit 18, the output control unit 19, the designation receiving unit 20, and the selection unit 21 are realized.

  In addition, the sales activity support system 37 includes a key designation receiving unit 38, a prediction unit 39, and a prediction output control unit 40.

  In the present embodiment, the user 12 operates the input device 25 and inputs designation of an incomplete business activity in order to predict the subsequent progress of the incomplete business activity.

  The key designation accepting unit 38 accepts designation of sales activity transition data relating to an incomplete sales activity that is input by the user 12 by operating the input device 25.

  Then, the key designation receiving unit 38 provides the prediction unit 39 with designation of incomplete sales activity transition data.

  When the prediction unit 39 receives designation of incomplete business activity transition data from the key designation accepting unit 38, the prediction unit 39 reads the business activity transition data 261 to 26m recorded in the recording unit 5.

  Further, the prediction unit 39 creates prediction data 41 indicating a prediction of a subsequent state in the incomplete business activity transition data based on the incomplete business activity transition data and other business activity transition data.

  Then, the prediction unit 39 records the prediction data 41 in the recording unit 5.

  FIG. 18 is a diagram illustrating an example of a method for predicting a subsequent progress state in incomplete business activity transition data.

  In normal time-series prediction, a current state and a past state are compared for a certain time-series change, and a state after the present is predicted based on the past state.

  In contrast, in the present embodiment, incomplete sales activity transition data 26b is used instead of the current state, and other sales activity transition data 26c to 26n are used instead of the past state, and other sales activity transition data is used. Based on 26c to 26n, the prediction data 41 that predicts the subsequent progress state of the incomplete sales activity transition data 26b is created.

  For example, as an example of the time series prediction technology, there is a time series short-term prediction technology by applying chaos (Shinji Murata, “Time series short-term prediction technology by applying chaos”, Meiden Times, Meidensha Corporation, April 25, 1995, No. 241, pages 49-56).

  The prediction output control unit 40 reads the prediction data 41 recorded in the recording unit 5, and based on the prediction data 41, the output device 23 predicts the incomplete sales activity transition data designated by the user 12. The process for displaying the progress of is performed.

  In the present embodiment described above, it is possible to check the progress of a predefined sales item or an item corresponding to a freely described expression. Subsequent progress can be predicted for items corresponding to the described expression. The user 12 can perform subsequent business activities with reference to this prediction.

(Fourth embodiment)
In the present embodiment, a modification of the third embodiment will be described.

  In this embodiment, sales activity transition data is modeled to create sales activity model data. Based on the sales activity model data, current status evaluation with predetermined sales activity transition data is performed, and sales activity support for predicting progress is provided. The system will be described.

  FIG. 19 is a block diagram showing an example of a sales activity support system according to the present embodiment.

  The computer 1 reads and executes the programs 2a and 2e recorded on the recording medium 53.

  The computer 1 operates in accordance with the program 2e and realizes a function as a sales activity support system 42.

  Similarly to the sales activity support system 4 according to the first embodiment, the sales activity support system 42 is a registration unit 13, a reading unit 14, a providing unit 15, an extraction item reception unit 16, a transition creation unit 17, an inconsistency. Functions as the determination unit 18, the output control unit 19, the designation receiving unit 20, and the selection unit 21 are realized.

  In addition, the sales activity support system 42 includes a classification unit 43, a modeling unit 44, a key designation receiving unit 45, an evaluation unit 46, and an evaluation output control unit 47.

  The classification unit 43 reads the business activity transition data 261 to 26m recorded in the recording unit, and classifies the business activity transition data 261 to 26m according to whether the result is “order received” or “missed”.

  Then, the classification unit 43 includes the sales activity transition data whose result is “order received” in the order classification data 48 and records the order classification data 48 in the recording unit 5.

  Further, the classification unit 43 includes the business activity transition data whose result is “missed” in the lost category data 49 and records the lost category data 49 in the recording unit 5.

  The modeling unit 44 reads the order classification data 48 recorded in the recording unit 5, performs time series mining on the order classification data 48 to create a time series model, and time series model data indicating the time series model 50 is recorded in the recording unit 5.

  FIG. 20 is a diagram illustrating an example of a time-series model.

  The modeling unit 44 reads out the missing classification data 49 recorded in the recording unit 5, performs time series mining on the missing classification data 49 to create a time series model, and shows a time series model. The model data 51 is recorded in the recording unit 5.

  The key designation accepting unit 45 accepts designation of sales activity transition data relating to a sales activity to be evaluated, which is input by the user 12 by operating the input device 25.

  Then, the key designation receiving unit 45 provides the evaluation unit 46 with designation of sales activity transition data to be evaluated.

  When the evaluation unit 46 receives the designation of the evaluation target business activity transition data from the key designation receiving unit 45, the evaluation unit 46 reads the evaluation target sales activity transition data recorded in the recording unit 5 and also reads the time-series model data 50. .

  The evaluation unit 46 also compares the sales activity transition data to be evaluated with the time-series model data 50, evaluates the sales activity transition data to be evaluated, and records evaluation data 52 indicating the evaluation result in the recording unit 5. .

  For example, when the evaluation target sales activity transition data is incomplete sales activity transition data, the evaluation unit 46 evaluates the current status of the evaluation target sales activity transition data compared to the time-series model data 50, and the progress prediction (of possible patterns). All are obtained), and the activity (at least one of the business item and expression) up to the preferable result “order received” among the prediction results is obtained, and the evaluation data 52 is created.

  In the present embodiment, the evaluation unit 46 performs the evaluation based on the time-series model data 50 when the result of “order received” is obtained. On the contrary, the result of “missed” is obtained. The evaluation may be performed based on the time series model data 51 obtained.

  The evaluation output control unit 47 reads the evaluation data 52 recorded in the recording unit 5, and based on the evaluation data 52, the evaluation result of the sales activity transition data to be evaluated designated by the user 12 is output to the output device 23. Process to display.

  For example, the invention disclosed in Japanese Patent Application Laid-Open No. 2002-351899 can be used for modeling the order classification data 48 and the unsuccessful classification data 49, and evaluation and prediction based on the business activity model.

  As a modeling technique, for example, a model self-organization technique can be used. As a method of model self-organization, for example, the document “Self-organization of Dialogue Task Model from Sample Word Time Series”, IEICE J83-D-II, No. 2, p. The extended image automaton self-organization method disclosed in 827-839 (2000) can be used.

  In the present embodiment described above, the progress state of the sales activity regarding the sales item defined in advance and the extracted item corresponding to the freely described expression is obtained.

  Therefore, it is possible to increase the information that the user 12 refers to in sales activities, and to support sales activities more effectively.

  Further, in the present embodiment, the evaluation target sales activity specified by the user 12 is compared with the sales activity model when the result of “order” is obtained, and an evaluation result based on the comparison is output.

  Therefore, when the evaluation target sales activity is completed, the user 12 can confirm the difference between the sales activity performed by the user 12 and the sales activity model for obtaining the result of “order received”.

  Further, when the evaluation target sales activity is incomplete, the user 12 can know what type of activity will increase the probability of obtaining the result of “order” from now on.

  In the present embodiment, the classification unit 43 classifies the sales activity transition data 261 to 26m based on “orders” and “missed”, but it is based on customers, customer industries, and their departments. In addition, sales activity transition data 261 to 26m may be classified. As described above, it is assumed that the classification unit 43 can classify the sales activity transition data 261 to 26m according to various standards.

(Fifth embodiment)
In this embodiment, a specific example of the second embodiment will be described.

  In the present embodiment, a sales activity support system for obtaining sales activity data related to a completed sales activity having a similar relationship with sales activity data related to an incomplete sales activity will be described. However, it goes without saying that the present embodiment is not limited to sales activity data, and may be applied to sales activity transition data as in the second embodiment.

  FIG. 21 is a block diagram illustrating an example of the sales activity support system according to the present embodiment. In addition to the configuration illustrated in FIG. 13, the feature amount extraction unit 61 that extracts feature amounts from the sales activity data 71 to 7n includes It has been added. Here, the sales activity data 71 to 7n are the same as those described with reference to FIG.

  FIG. 22 is a schematic diagram illustrating an example of processing by the sales activity support system according to the present embodiment.

  The feature amount extraction unit 61 converts each completed business activity data 71 to 7n-1 into feature amount data 621 to 625.

  Here, the feature quantity data 621 to 625 are k-dimensional real vectors expressed in the form of (w1, w2,..., Wk). The feature amount component wi is a feature amount related to the feature amount item ti. The feature amount item ti may be, for example, a word obtained by text mining from free text data. Or it is good also as each item of sales item data. Or it is good also as each item of sales activity identification data. Alternatively, a combination of these may be used.

  The feature amount represents the importance of the feature amount item in the business activity data, and may be, for example, the relative appearance frequency of text data. Alternatively, the value of tfidf described in the following document [S1] may be used.

Reference [S1] “Machine Learning in Automated Text Categorization”, Fabrizio Sebastiani, ACM Computing Surveys, Vol. 34, No. 1, March 2002, pp. 1-47
Further, in the case of the feature amount related to the sales item data 10, it may be 1 when each item of the sales item data 10 is selected, and may be 0 when the item is not selected. In addition, factor analysis may be performed on a plurality of items, and the factor score may be used as a feature amount.

  In this way, feature amount data 621 to 625 can be obtained for each completed business activity data 71 to 7n-1.

  The feature quantity extraction unit 61 obtains feature quantity data 626 for the incomplete sales activity data 7n as well as the completed sales activity data.

  Next, the similarity determination unit 32 classifies the feature data 621 to 625 and divides them into categories. As the classification category, data existing in each completed sales activity data 71 to 7n−1 is used. For example, “order received” or “missed” in the business item data 10 may be used. When the feature amount data 621 to 625 of the present embodiment is used, the text classification method described in the document [S1] can be used for classification. For example, the feature quantity data 621 to 625 related to the completed sales activity data 71 to 7n-1 are used as teacher data, and the feature quantity data 626 related to the incomplete sales activity data 7n is classified, so that the category of the incomplete sales activity data 7n is determined. can get.

Hereinafter, the operation of the present embodiment will be described in detail using specific examples.
Now, it is assumed that the sales activity data is specified by the key specification receiving unit 31 and the sales activity data 71 to 71 n are read from the recording unit 5 by the similarity determination unit 32.

  FIG. 23 is an example of the input business activity data 71 to 7n. In this example, attention is paid to free text data, and therefore other data items are not shown.

  Here, no. 1 to 5 are examples of completed business activity data 71 to 7n-1 (where n = 5). Data 6 is an example of incomplete sales activity data 7n.

  Next, the feature quantity extraction unit 61 extracts feature quantities from these completed or incomplete sales activity data 71 to 7n.

  As features, the expressions included in the free text data are "Demonstration is well received", "Positive questions are issued", "Other companies withdraw", "Demonstration is unpopular", and "Slow response" The following five are adopted.

  As the feature quantity, the appearance frequency of each feature quantity item is taken. Therefore, the feature amount of each sales activity data is a value shown in FIG. In this example, the norm is standardized to be 1. No. Since the sales activity data 6 includes the expressions 1 and 2 in FIG. 23, w1 and w2 of the feature amounts are set to 0.7 (= 1 / √2), respectively.

  Next, the similarity determination unit 32 performs classification. The categories of the sales activity data 71 to 7n are shown in FIG. Here, as an example of the classification method, an on-line method described in the document [S1] is used. Instead of the online method, for example, the following methods (i) to (v) described in the document [S1] may be used.

(I) Naive Bayes classifier.
(Ii) Decision tree.
(Iii) Regression analysis.
(Iv) Neural network.
(V) support vector machine.
As a result of the online method, the classifier in the data of this embodiment has the values shown in FIG.

  The online method is a method for obtaining a classifier belonging to each category, and setting the value of the inner product of the feature quantity data and the classifier as the degree to which the feature quantity data belongs to the category. In the present embodiment, when the classifier is obtained, a perceptron algorithm is used, and the learning rate is set to 0.3 and the norm is normalized to 1.

  Finally, the similarity determination unit 32 classifies the incomplete business activity data 7n. When the inner product of the classifier of each category and the feature amount data is obtained, the order is 0.742 and the case of lost is 0.448. In this example, it is classified as an order.

  Therefore, no. According to the incomplete sales activity data 7n having the feature quantity of 6, it is expected that the possibility of ordering is high.

  Here, when the category is, for example, “order received” or “missed” in the business item data 10, it is predicted whether the uncompleted sales activity data 7n is likely to be “order received” or “missed”. Can be used by managers to manage sales activities.

  Moreover, since the feature data having similar inner products has high similarity, the feature data of the completed business activity data 71 to 73 having high similarity to the feature data related to the unfinished business activity data 7n is obtained. be able to. As a result, similar sales activity data can be obtained, and each sales person can refer to future sales activities.

  Subsequently, the similarity determination unit 32 selects completed sales activity data corresponding to the obtained feature data having a high degree of similarity, and creates similarity determination data 34 indicating the selection result.

  Then, the similarity determination unit 32 records the similarity determination data 34 in the recording unit 5. The similarity determination unit 32 may output the category classification result in parallel with the recording of the similarity determination data 34 or without creating the similarity determination data 34.

  As described above, according to the present embodiment, in addition to the effects of the second embodiment, feature amount data is extracted from incomplete business activity data, and based on the classification result of each completed business activity data, By classifying feature quantity data related to sales activity data into categories and outputting a classification result of incomplete sales activity data, it is possible to classify sales activities in progress in the same way as completed sales activities.

  In addition, since it is possible to predict the possibility that incomplete sales activity data will be “order received” or “missed”, the administrator can use it for managing sales activities. Furthermore, each sales representative can use similar sales activity data as a reference for future sales activities.

  The programs 2a to 2e according to the above embodiments are written in recording media 2, 29, 36, 53 such as a magnetic disk (flexible disk, hard disk, etc.), an optical disk (CD-ROM, DVD, etc.), and a semiconductor memory, respectively. Therefore, it can be applied to the computer 1.

  Further, the programs 2a to 2e can be transmitted through a communication medium and applied to the computer 1, respectively. The computer 1 reads the programs 2a to 2e, respectively, and realizes the above functions by controlling operations by the programs 2a to 2e.

  In each of the above embodiments, the function realized by software may be realized by hardware.

  In addition, the components of the text mining system 3 and the sales activity support systems 4, 30, 37, and 42 according to the above embodiments may be rearranged as long as the same action can be realized. The constituent elements may be freely combined, and any of the constituent elements may be deleted. For example, the recording unit 5 may be divided into a plurality of parts.

  Further, in the text mining system 3 and the sales activity support systems 4, 30, 37, and 42 according to the present embodiment, the programs 2a to 2e are distributed and arranged on a plurality of computers, and the processes are executed in cooperation with each other. It may be.

  The present invention is effective in the field of supporting sales activities based on stored data.

The block diagram which shows an example of the sales activity support system which concerns on the 1st Embodiment of this invention. The figure which shows an example of the sales activity data which concerns on the embodiment. The figure which shows an example of dictionary data. The figure which shows an example of the screen for business activity data registration. The figure which shows an example of the production state of the sales activity transition data by a transition production part. The figure which shows an example of the method of the consistency judgment by an inconsistency judgment part. The figure which shows an example of the output result by a transition output part. The figure which shows the 1st example of the output result by a mismatching output part. The figure which shows the 2nd example of the output result by a mismatching output part. The figure which shows an example of the output result by a detailed output part. The flowchart which shows an example of the first half part of the process by the sales activity support system and text mining system which concern on the embodiment. The flowchart which shows an example of the latter half part of the process by the business activity support system and text mining system which concern on the embodiment. The block diagram which shows an example of the sales activity support system which concerns on the 2nd Embodiment of this invention. The figure which shows an example of the judgment method of the similarity between unfinished sales activity transition data and completed sales activity transition data. The figure which shows an example of the display of similarity judgment data. The flowchart which shows an example of the similarity determination process by the sales activity support system which concerns on the embodiment. The block diagram which shows an example of the sales activity support system which concerns on the 3rd Embodiment of this invention. The figure which shows an example of the prediction method of the subsequent progress state in incomplete sales activity transition data. The block diagram which shows an example of the sales activity support system which concerns on the 4th Embodiment of this invention. The figure which shows an example of a time series model. The block diagram which shows an example of the sales activity support system which concerns on the 5th Embodiment of this invention. The schematic diagram which shows an example from extraction of feature-value data to a similarity determination process. The figure which shows an example of free text data. The figure which shows an example of feature-value data. The figure which shows an example of a category. The figure which shows an example of a classifier.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... Computer, 2, 29, 36, 53 ... Recording medium, 2a-2e ... Program, 3 ... Text mining system, 3a ... Data input part, 3b ... Analysis part, 3c ... Data output part, 4, 30, 37, 42 ... Sales activity support system, 5 ... Recording unit, 6 ... Dictionary data, 71-7n ... Sales activity data, 13 ... Registration unit, 14 ... Reading unit, 15 ... Providing unit, 16 ... Extracted item receiving unit, 17 ... Transition Creation unit, 18 ... mismatch judgment unit, 19 ... output control unit, 19a ... mismatch output unit, 19b ... transition output unit, 19c ... detailed output unit, 20 ... designation reception unit, 21 ... selection unit, 31, 38, 45 ... Key designation acceptance unit, 32 ... Similarity determination unit, 33 ... Similar output control unit, 34 ... Similarity determination data, 39 ... Prediction unit, 40 ... Predictive output control unit, 41 ... Prediction data, 43 ... Classification unit, 44 ... Modeling part, 46 ... Valuation part, 47 ... Evaluation output control part, 48 ... Order classification data, 49 ... Missed classification data, 50, 51 ... Time series model data, 52 ... Evaluation data, 61 ... Feature quantity extraction part, 621 to 626 ... Feature quantity data.

Claims (15)

On the computer,
At least one of sales item data indicating any of the business items defined in advance as being implemented in the sales activity, free text data freely input with respect to the sales activity, time data, and A function for reading a plurality of sales activity data associated with sales activity identification data for identifying a sales activity from a recording means;
When the free text data is included in any of the plurality of sales activity data, the free text mining means extracts a predetermined item registered in the dictionary data in advance based on the text data. The ability to provide text data,
A function of accepting an extraction item extracted from the free text data by the text mining means;
A program for realizing a function of creating sales activity transition data representing the sales items and the extracted items in a time series based on the time data using sales activity data having the same sales activity identification data. The program according to claim 1,
On the computer,
In the sales activity transition data, a function for determining whether or not the content of the business item and the content of the extracted item are consistent,
A program for further realizing a mismatch output function for outputting a message to the effect that it is determined by the determination function that there is no matching. The program according to claim 2,
The sales activity identification data includes person identification data for identifying a sales person or department data for identifying a sales department.
The inconsistency output function outputs, for each person-in-charge identification data or the department data, an aggregation result when it is determined by the determination function that there is no consistency. In the program according to any one of claims 1 to 3,
On the computer,
Transition output function for outputting the sales activity transition data as a graph,
A function of accepting designation of an arbitrary position in the graph;
A function of selecting the sales activity data corresponding to the arbitrary position of the sales activity transition data;
A program for further realizing a detailed output function for outputting sales activity data selected by the selection function. On the computer,
At least one of sales item data indicating any of the business items defined in advance as being implemented in the sales activity, free text data freely input with respect to the sales activity, time data, and A function for reading a plurality of sales activity data associated with sales activity identification data for identifying a sales activity from a recording means;
When the free text data is included in any of the plurality of sales activity data, the free text mining means extracts a predetermined item registered in the dictionary data in advance based on the text data. The ability to provide text data,
A function of accepting an extraction item extracted from the free text data by the text mining means;
A function of creating a plurality of sales activity transition data representing the sales items and the extracted items in a time series based on the time data, using the sales activity data having the same sales activity identification data;
A function of extracting other sales activity transition data similar to predetermined sales activity transition data for the plurality of sales activity transition data;
A program for realizing a similar output function that outputs extracted business activity transition data. The program according to claim 5,
The program is characterized in that the extraction function extracts business activity transition data similar to predetermined business activity transition data based on a result of a vector operation performed on the plurality of business activity transition data. On the computer,
At least one of sales item data indicating any of the business items defined in advance as being implemented in the sales activity, free text data freely input with respect to the sales activity, time data, and A function for reading a plurality of sales activity data associated with sales activity identification data for identifying a sales activity from a recording means;
When the free text data is included in any of the plurality of sales activity data, the free text mining means extracts a predetermined item registered in the dictionary data in advance based on the text data. The ability to provide text data,
A function of accepting an extraction item extracted from the free text data by the text mining means;
A function of creating a plurality of sales activity transition data representing the sales items and the extracted items in a time series based on the time data, using the sales activity data having the same sales activity identification data;
A program for realizing a function of creating a model indicating a transition relationship between the sales item and the extracted item based on the plurality of sales activity transition data. The program according to claim 7, wherein
On the computer,
A program characterized by applying the model to further realize a function of predicting at least one of a business item and an extraction item that occur later with respect to business activity transition data whose sales activity has not ended. The program according to claim 8, wherein
The forecast function is configured to select a business item that is set as a target when multiple types of business items and extracted items that are generated later are predicted for business activity transition data that has not been terminated. A program characterized by extracting predictions including. At least one of sales item data indicating any of the business items defined in advance as being implemented in the sales activity, free text data freely input with respect to the sales activity, time data, and Means for reading a plurality of sales activity data associated with the sales activity identification data for identifying the sales activity from the recording means;
When the free text data is included in any of the plurality of sales activity data, the free text mining means extracts a predetermined item registered in the dictionary data in advance based on the text data. Means for providing text data;
Means for receiving an extraction item extracted from the free text data by the text mining means;
A sales activity support system comprising means for creating sales activity transition data representing the sales items and the extracted items in a time series based on the time data using the sales activity data having the same sales activity identification data . In a method of supporting sales activities by computer,
At least one of sales item data indicating any of the business items defined in advance as being implemented in the sales activity, free text data freely input with respect to the sales activity, time data, and A plurality of sales activity data related to the sales activity identification data for identifying the sales activity is read from the recording means,
When the free text data is included in any of the plurality of sales activity data, the free text mining means extracts a predetermined item registered in the dictionary data in advance based on the text data. Provide text data,
Accepting an extraction item extracted from the free text data by the text mining means;
A method of creating sales activity transition data representing the sales items and the extracted items in a time series based on the time data, using the sales activity data having the same sales activity identification data. On the computer,
At least one of sales item data indicating any of the business items defined in advance as being implemented in the sales activity, free text data freely input with respect to the sales activity, time data, and A function for reading a plurality of sales activity data associated with sales activity identification data for identifying a sales activity from a recording means;
When the free text data is included in any of the plurality of sales activity data, the free text mining means extracts a predetermined item registered in the dictionary data in advance based on the text data. The ability to provide text data,
A function of accepting an extraction item extracted from the free text data by the text mining means;
A function of extracting a plurality of feature amount data from a plurality of completed sales activity data based on the sales item and the extracted item, using the sales activity data having the same sales activity identification data;
A function of classifying the plurality of feature amount data into categories represented by business item data included in each completed business activity data;
A function of extracting feature amount data from incomplete sales activity data based on the sales item and the extracted item, using the sales activity data having the same sales activity identification data;
A function of classifying feature quantity data related to the incomplete sales activity data into categories based on a classification result of each of the completed sales activity data;
A program for realizing a classification result output function for outputting a classification result of the incomplete sales activity data.   The category represents a received order or a lost order. In the program according to claim 12 or 13,
The classification function is a program that executes the classification based on a result of a vector operation performed on the feature amount data. The program according to any one of claims 12 to 14,
In the computer,
A function of extracting feature quantity data similar to feature quantity data relating to the incomplete sales activity data from among a plurality of feature quantity data relating to the completed sales activity data;
A program for further realizing a similar output function for outputting sales activity data corresponding to the similar feature data.

Images