WO2024134795A1 - 情報処理装置、分析支援方法、および分析支援プログラム - Google Patents

情報処理装置、分析支援方法、および分析支援プログラム Download PDF

Info

Publication number
WO2024134795A1
WO2024134795A1 PCT/JP2022/047046 JP2022047046W WO2024134795A1 WO 2024134795 A1 WO2024134795 A1 WO 2024134795A1 JP 2022047046 W JP2022047046 W JP 2022047046W WO 2024134795 A1 WO2024134795 A1 WO 2024134795A1
Authority
WO
WIPO (PCT)
Prior art keywords
path
paths
elements
information processing
display control
Prior art date
Legal status (The legal status 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 status listed.)
Ceased
Application number
PCT/JP2022/047046
Other languages
English (en)
French (fr)
Japanese (ja)
Inventor
克也 速水
昌紀 小泉
健太 笠原
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
NEC Corp
Original Assignee
NEC Corp
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 NEC Corp filed Critical NEC Corp
Priority to JP2024565468A priority Critical patent/JPWO2024134795A1/ja
Priority to PCT/JP2022/047046 priority patent/WO2024134795A1/ja
Publication of WO2024134795A1 publication Critical patent/WO2024134795A1/ja
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N7/00—Computing arrangements based on specific mathematical models
    • G06N7/01—Probabilistic graphical models, e.g. probabilistic networks

Definitions

  • the present invention relates to an information processing device that performs processing related to causal analysis.
  • Patent Document 1 describes constructing a directed acyclic graph that represents the relationships between multiple variables and a target outcome. The same document also describes using the constructed directed acyclic graph to analyze whether there is sufficient causal evidence to identify a causal relationship between a variable of interest and the target outcome.
  • a directed acyclic graph represents each element as a node and shows the causal relationships between them as directed edges, and although it is intuitively easy to understand, as the number of elements increases, its appearance becomes cluttered. Furthermore, it is difficult for users to quickly recognize the information they want to know from a cluttered directed acyclic graph.
  • One aspect of the present invention aims to realize an information processing device or the like that can allow a user to easily recognize the results of causal analysis.
  • An information processing device includes a receiving means for receiving the designation of at least one element from among a plurality of elements that have been causally analyzed, a path detection means for detecting a path that includes the designated element from among paths made up of a plurality of the elements that are causally linked, and a display control means for displaying path information indicating each path in a manner corresponding to each path when the path detection means detects a plurality of paths that share the element that serves as an end point.
  • An analysis support method includes at least one processor receiving the designation of at least one element from among a plurality of elements that have been causally analyzed, detecting a path that includes the designated element from among paths made up of the plurality of elements that are causally linked, and, when a plurality of paths that share the common end point element are detected, displaying path information indicating each path in a manner appropriate to each path.
  • An analysis support program causes a computer to function as: a receiving means for receiving the designation of at least one element among a plurality of elements that have been causally analyzed; a path detection means for detecting a path that includes the designated element among paths made up of a plurality of the elements that are causally linked; and a display control means for displaying path information indicating each path in a manner corresponding to each path when the path detection means detects a plurality of paths that share the element that serves as the end point.
  • FIG. 1 is a block diagram showing a configuration of an information processing device according to a first exemplary embodiment of the present invention
  • 1 is a flow chart showing the flow of an analysis support method according to a first exemplary embodiment of the present invention.
  • FIG. 11 is a block diagram showing a configuration of an information processing device according to an exemplary embodiment 2 of the present invention.
  • 13A and 13B are diagrams illustrating an example of element specification and path detection.
  • FIG. 13 is a diagram showing an example of a display screen for path information.
  • FIG. 13 is a diagram showing another example of the path information display screen.
  • FIG. 13 is a diagram showing yet another example of the path information display screen.
  • FIG. 13 is a diagram showing yet another example of the path information display screen.
  • FIG. 13 is a diagram showing yet another example of the path information display screen.
  • FIG. 11 is a flow chart showing the flow of an analysis support method according to an exemplary embodiment 2 of the present invention.
  • FIG. 1 is a diagram showing an example of a computer that executes instructions of a program, which is software that realizes the functions of each device according to each exemplary embodiment of the present invention.
  • Example embodiment 1 DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
  • This exemplary embodiment is a basic form of the exemplary embodiments described below.
  • Fig. 1 is a block diagram showing the configuration of the information processing device 1. As shown in the figure, the information processing device 1 includes a reception unit 11, a path detection unit 12, and a display control unit 13.
  • the reception unit 11 receives the designation of at least one element out of the multiple elements that have been subjected to causal analysis.
  • the path detection unit 12 detects a path that includes a specified element from among paths consisting of multiple elements that are causally connected.
  • the display control unit 13 displays path information indicating each path in a manner appropriate to each path.
  • the information processing device 1 is configured to include a receiving unit 11 that receives the designation of at least one element among the multiple elements that have been causally analyzed, a path detection unit 12 that detects paths that include the specified element among paths consisting of the multiple elements that are causally connected, and a display control unit 13 that displays path information indicating each path in a manner corresponding to each path when the path detection unit 12 detects multiple paths that share a common end element. Therefore, the information processing device 1 according to this exemplary embodiment has the effect of allowing the user to more easily recognize the analysis results of the causal analysis.
  • the functions of the information processing device 1 described above can also be realized by a program.
  • the analysis support program according to this exemplary embodiment is configured to cause a computer to function as a receiving means for receiving the designation of at least one element among the elements that have been causally analyzed, a path detection means for detecting a path including the designated element among the paths consisting of the elements that are causally connected, and a display control means for displaying path information indicating each path in a form corresponding to each path when the path detection means detects multiple paths that have a common end point element. Therefore, the analysis support program according to this exemplary embodiment has the effect of allowing a user to easily recognize the analysis result of the causal analysis.
  • Flow of analysis support method The flow of the analysis support method according to this exemplary embodiment will be described with reference to Fig. 2.
  • Fig. 2 is a flow diagram showing the flow of the analysis support method. Note that the execution subject of each step in this analysis support method may be a processor provided in the information processing device 1, or a processor provided in another device, or each step may be a processor provided in a different device.
  • At least one processor accepts the designation of at least one element among the multiple elements that have been causally analyzed.
  • At least one processor detects a path that includes the element specified in S11 from among paths consisting of multiple elements that are causally connected.
  • At least one processor displays path information indicating the paths detected in S12. If multiple paths that share a common end element are detected in S12, in S13, the processor displays path information indicating each path in a manner appropriate to each path.
  • the analysis support method employs a configuration in which at least one processor receives the designation of at least one element among the multiple elements that have been causally analyzed, detects paths that include the specified element among paths consisting of the multiple elements that are causally connected, and, when multiple paths that share a common end element are detected, displays path information indicating each path in a manner appropriate to each path. Therefore, the analysis support method according to this exemplary embodiment has the effect of allowing the user to more easily recognize the results of the causal analysis.
  • Fig. 3 is a block diagram showing the configuration of the information processing device 2.
  • the information processing device 2 includes a control unit 20 that controls each unit of the information processing device 2, and a storage unit 21 that stores various data used by the information processing device 2.
  • the information processing device 2 also includes a communication unit 22 that allows the information processing device 2 to communicate with other devices, an input unit 23 that accepts input of various data to the information processing device 2, and a display unit 24 that allows the information processing device 2 to display and output various data. Note that each of the components from the storage unit 21 to the display unit 24 may be built into the information processing device 2, or may be an external device attached to the information processing device 2.
  • the control unit 20 also includes a reception unit 201, a causal analysis unit 202, a path detection unit 203, a change index calculation unit 204, a path coefficient calculation unit 205, a condition satisfaction determination unit 206, a cost calculation unit 207, a path information generation unit 208, and a display control unit 209.
  • the reception unit 201 receives the designation of at least one element from among the multiple elements causally analyzed by the causal analysis unit 202. Note that the term “element” is explained in the next paragraph, and the designation of elements is explained later in the section “Designation of elements and detection of paths.”
  • the causal analysis unit 202 performs causal analysis on multiple given elements, and identifies the causal structure and causal relationships between those elements.
  • the elements that are the subject of the causal analysis may be each question in a questionnaire and the answers to those questions, or the elements that are the subject of the causal analysis may be the actions and their results shown in the behavioral history information of a subject or group of subjects.
  • the data format of each element may or may not be standardized.
  • the causal analysis unit 202 may perform the causal analysis by applying known causal inference or causal search methods that correspond to the target elements, etc.
  • each element is a variable related to a specific event.
  • a variable related to the event of a product being purchased e.g., a numerical value indicating the number of products purchased or the sales price
  • a variable indicating the price of a product, which is related to the above event can also be an element.
  • the information processing device 2 does not necessarily need to include the causal analysis unit 202. If the information processing device 2 does not include the causal analysis unit 202, it is sufficient for the information processing device 2 to acquire the results of the causal analysis performed in another device via the communication unit 22 or the input unit 23.
  • the path detection unit 203 detects paths that include a specified element from among paths consisting of multiple elements that are causally related.
  • a "path” can also be referred to as a group of elements or a string of elements consisting of multiple elements that are causally related. Note that the element specification is accepted by the acceptance unit 201 as described above. Details of the path detection method will be explained later in the section "Specifying elements and detecting paths.”
  • each element that has undergone causal analysis can be represented in a causal graph.
  • each element is represented as a "node,” and the causal relationships between each element are represented as "edges" connecting the nodes.
  • nodes are classified into RCVs (Root Cause Variables), which are the start points of paths, targets, which are the end points of paths, and variables located between the RCVs and targets.
  • RCVs Root Cause Variables
  • a "path” can also be expressed as a route consisting of one or more edges connecting an RCV and a target, or as a matrix of nodes from an RCV to a target that are connected by edges.
  • a "path” can be classified into direct paths, which directly connect an RCV and a target, and indirect paths, which involve one or more variables between the RCV and the target.
  • the target which is the end point of the path, is sometimes called the objective variable, response variable, reaction variable, outcome variable, dependent variable, or non-explanatory variable.
  • the RCV which is the start point of the path, and the variables located between the RCV and the target are sometimes called explanatory variables, predictor variables, or independent variables.
  • the change index calculation unit 204 calculates an index value that indicates the degree to which the element at the end of a path, i.e., the target variable, changes when the value of the element at the start of the path detected by the path detection unit 203, i.e., the RCV, is changed. This process may be performed for each of the multiple paths detected by the path detection unit 203.
  • the index value indicating the degree of change may indicate, for example, the amount of change or the rate of change.
  • the change index calculation unit 204 may calculate the amount of change in the target variable when the RCV is increased by a predetermined amount (for example, 1) as the index value.
  • This index value can be calculated by multiplying the amount of change in the RCV by a path coefficient, which will be described later. Note that, when there are multiple paths, the index value can be calculated by the sum of the products of the amount of change in the RCV and each path coefficient.
  • the index value may also indicate a change in a statistical quantity.
  • the change index calculation unit 204 may calculate the index value indicating a change in a statistical quantity of the target variable when the RCV is increased by a predetermined amount (for example, 1).
  • the statistical quantity may be, for example, a standard deviation.
  • the change index calculation unit 204 may calculate, as the index value, the ratio of the standard deviation of the target variable after the RCV is changed to the standard deviation of the target variable before the RCV is changed.
  • the user may also be able to specify the statistical quantity to be targeted.
  • the path coefficient calculation unit 205 calculates a path coefficient for each path detected by the path detection unit 203.
  • the path coefficient is a coefficient calculated using the weight of each edge connecting each element that has been causally analyzed. Specifically, for a direct path, the path coefficient calculation unit 205 calculates the weight of the edges that make up the path. For an indirect path, the path coefficient calculation unit 205 calculates the product of the weights of the multiple edges that make up the path. Note that the product of the edge weights is merely one example of a path coefficient, and path coefficients may be calculated using other methods. It can be said that a path with a larger path coefficient value is more likely to be a useful path.
  • the condition satisfaction determination unit 206 determines whether or not each path detected by the path detection unit 203 satisfies a predetermined condition. Any condition can be set, and the user may be allowed to set the conditions. For example, the condition may be that a specific element is included, or a condition may be set regarding the index value calculated by the change index calculation unit 204, the path coefficient calculated by the path coefficient calculation unit 205, or the cost calculated by the cost calculation unit 207. For example, the condition satisfaction determination unit 206 may determine that a path whose cost calculated by the cost calculation unit 207 is equal to or less than a predetermined threshold satisfies the condition, and that a path whose cost is less than the threshold does not satisfy the condition.
  • the cost calculation unit 207 calculates the cost required to change the value of the element at the end of each path detected by the path detection unit 203 by a specified amount (the target variable described above). Note that "cost" includes at least one of monetary cost, time cost, and human cost. The cost calculation unit 207 may also calculate amounts of money, time, number of people, etc., or may calculate index values indicating the amount of these. This allows the cost calculation unit 207 to calculate the cost based on the relationship.
  • the cost calculation unit 207 may first cause the change index calculation unit 204 to calculate the amount of change in the value of another element that is required to change the value of the target variable by a predetermined amount. Then, the cost calculation unit 207 may calculate the cost required to change the value of the target variable by the calculated amount of change. Note that the cost required to change the value of each variable by a unit amount may be specified in advance.
  • the cost calculation unit 207 may, for example, calculate the cost when each element contained in the path is changed. The cost calculation unit 207 may then determine the average value of the calculated costs, or the minimum or maximum value of the calculated costs, as the cost of the path.
  • the path information generating unit 208 generates path information indicating each path detected by the path detection unit 203. In addition, when the path detection unit 203 detects multiple paths that share a common end point element, the path information generating unit 208 generates path information indicating each path in a manner appropriate to each path.
  • the display control unit 209 causes the display unit 24 to display the path information generated by the path information generation unit 208. As described above, this path information indicates the paths detected by the path detection unit 203. Furthermore, when the path detection unit 203 detects multiple paths that share a common end element, the display control unit 209 causes the display unit 209 to display path information indicating each path in a manner appropriate to each path.
  • Fig. 4 is a diagram showing an example of element designation and path detection. Note that in Fig. 4, elements subjected to causal analysis are shown as circular nodes, and causal relationships between the elements are shown as directed edges (one-way arrows). Fig. 4 also shows five examples, examples 1 to 5.
  • the element of node N4 is specified as the start point of the path (the RCV described above), and the element of node N1 is specified as the end point of the path (the target described above).
  • the receiving unit 201 may receive the specification of both the element that is the start point and the element that is the end point of the path.
  • the path detection unit 203 detects a path that starts from the element of node N4 and ends at the element of node N1, and that connects the elements through a causal relationship. Specifically, the path detection unit 203 detects a path from the start element to the end element by tracing the directed edges along their directions. If there are branches, the path detection unit 203 detects each branch as a separate path. In Example 1, a path p1 that runs from the element of node N4 to the element of node N1 via the element of node N2, and a path p2 that runs from the element of node N4 to the element of node N1 via the element of node N3 are detected.
  • Example 2 is an example in which the element of node N4 is specified as the start point of the path.
  • the receiving unit 201 may receive the specification only for the element that is to be the start point of the path.
  • the path detection unit 203 detects a path in which the elements are causally connected, with the element of node N4 as the start point. Specifically, in example 2, the same paths p1 and p2 as in example 1 are detected.
  • an element of node N4 and an element of node N5 are specified as the start point of a path, and an element of node N1 is specified as the end point of the path.
  • the receiving unit 201 may receive the specification of multiple elements as the start point of a path.
  • the path detection unit 203 detects a path in which elements are causally connected with the element of node N4 as the start point and the element of node N1 as the end point, and a path in which elements are causally connected with the element of node N5 as the start point and the element of node N1 as the end point.
  • a path p1 from the element of node N4 to the element of node N1 via the element of node N2, and a path p2 from the element of node N5 to the element of node N1 via the element of node N3 are detected.
  • Example 4 is an example in which the element of node N1 is specified as the end point of the path.
  • the receiving unit 201 may only receive the specification of the element that is the end point of the path.
  • the path detection unit 203 detects a path in which the elements are causally connected, with the element of node N1 as the end point. Specifically, the path detection unit 203 detects a path from the end element to the start element by tracing the directed edges backwards.
  • a path p1 from the element of node N4 to the element of node N1 via the element of node N2, and a path p2 from the element of node N5 to the element of node N1 via the element of node N3 are detected.
  • the element of node N4 is specified as an element to be included in the path
  • the element of node N1 is specified as the end point of the path.
  • the receiving unit 201 may receive the specification of the elements to be included in the path.
  • the path detection unit 203 detects a path that includes the element of node N4, ends at the element of node N1, and has each element connected by a causal relationship. Specifically, the path detection unit 203 performs a process of tracing the directed edge backward from the specified element and a process of tracing along the direction of the directed edge, and detects a path that passes through the specified element and reaches the end element.
  • Example 5 a path p1 is detected that goes from the element of node N5 to the element of node N1 via the element of node N4 and the element of node N2, and a path p2 is detected that goes from the element of node N5 to the element of node N1 via the element of node N4 and the element of node N3.
  • the receiving unit 201 may also receive the designation of an indirect element.
  • the receiving unit 201 may receive the designation of a condition that an element to be included in a path must satisfy.
  • the path detection unit 203 extracts elements that satisfy the specified condition, and detects paths that include these elements. Examples of conditions include being (or not being) a direct or indirect cause or result of a specified node, being located upstream (or downstream) of one or more specified nodes, and combinations of these.
  • Fig. 5 is a diagram showing an example of a display screen of path information. More specifically, Fig. 5 shows an example of a display screen in which path information is displayed on a causal graph (also called a directed acyclic graph) in which a plurality of elements subjected to causal analysis are regarded as nodes and the causal relationships between the elements are indicated by directed edges (one-way arrows). Note that instead of a directed acyclic graph, a CPDAG (Completed Partially Directed Acyclic Graph) or an undirected graph may be displayed.
  • a causal graph also called a directed acyclic graph
  • CPDAG Completed Partially Directed Acyclic Graph
  • the causal graph shown in Figure 5 contains nodes N1 to N8. Each node is connected by a directed edge. Of the two nodes connected by a directed edge, the upstream side (the side at the base of the arrow) indicates the cause, and the downstream side (the side at the tip of the arrow) indicates the result. For example, nodes N6 and N7 are connected by a directed edge going from node N6 to N7. This shows that the event "purchasing a product", which is an element of node N6, causes the event "granting preferential treatment", which is an element of node N7.
  • the display screen shown in FIG. 5 also includes the causal graph described above, as well as objects a1 to a3 for accepting user input.
  • Object a1 is an object for accepting the designation of an element that is to be the end point of the path (the target described above).
  • the user selects object a1 and designates the element that is to be the end point by performing an input operation via the input unit 23.
  • the display control unit 209 may display a list of elements that can be used as end points among the elements included in the causal graph, and allow the user to designate the element that is to be used as the end point from among the elements displayed in the list.
  • a character string indicating the designated element is displayed on object a1, as shown in the figure.
  • Object a2 is an object for accepting the designation of an element to be the starting point of the path (the RCV described above).
  • the user selects object a2 and designates the element to be used as the starting point by performing an input operation via the input unit 23.
  • the display control unit 209 may display a list of elements included in the causal graph that can be used as the starting point, and allow the user to designate the element to be used as the starting point from among the elements displayed in the list.
  • a character string indicating the designated element is displayed on object a2, as shown in the figure.
  • Object a3 is an object for adding an element to be used as a starting point.
  • the display control unit 209 may display a list of nodes that can be used as starting points among the elements included in the causal graph, and allow the user to specify an element to be used as a starting point from the listed elements.
  • a new object indicating the specified element is displayed.
  • the path detection unit 203 detects paths that include the specified nodes.
  • "Product Purchase” an element of node N6, is specified as the end point
  • "Promotion” an element of node N2 is specified as the start point.
  • the path detection unit 203 detects the path "Node N2-Node N4-Node N6" and the path "Node N2-Node N5-Node N6".
  • the path information generation unit 208 When multiple paths are detected in this way, the path information generation unit 208 generates path information indicating each path in a manner appropriate to each path, and the display control unit 209 displays this.
  • the display control unit 209 displays the border of the end node in the detected path as a double line and the border of the start node as a thick line, and also displays the arrows between the nodes included in the path as thick lines. These displays show the detected paths. That is, in Figure 5, the nodes and directed edges displayed in this manner constitute the path information.
  • the line thickness of the directed edge in "Node N2-Node N4-Node N6" is different from that of the directed edge in "Node N2-Node N5-Node N6". This difference reflects the difference in the detected paths. In other words, by changing the line thickness of the directed edges in Figure 5, the path information is displayed in a manner that corresponds to the path.
  • the display control unit 209 may display path information on a causal graph in which multiple elements that have been causally analyzed are treated as nodes and the causal relationships between each element are shown as directed edges.
  • this has the effect of allowing the user to recognize the relationship between the causal graph and the detected path information, that is, the position of the detected path in the causal graph.
  • the criteria for determining the display mode of the path information of each path are not particularly limited.
  • the display mode may be determined based on at least any of the calculation results or judgment results of the change index calculation unit 204, the path coefficient calculation unit 205, the condition satisfaction judgment unit 206, and the cost calculation unit 207.
  • the display format of the path information is not particularly limited as long as it allows each path to be identified.
  • each path may be identified based on the shape of the directed edge (color, thickness, arrowhead size, etc.), or based on the display format of the node (color, outline, size, etc.).
  • the information processing device 2 includes a change index calculation unit 204 that calculates an index value indicating the degree of change in a variable at the end point of a path when a variable at the start point of the path is changed, for each of a plurality of paths detected by the path detection unit 203. Therefore, the path information generation unit 208 can generate path information in a manner corresponding to the calculated index value, and the display control unit 209 can display the path information in a manner corresponding to the calculated index value.
  • the effect of allowing the user to easily recognize the difference in the degree of change in the variable of the element at the end of each path when the variable of the element at the start of the path is changed can be obtained.
  • the information processing device 2 also includes a path coefficient calculation unit 205 that calculates a path coefficient for each of the multiple paths detected by the path detection unit 203. Therefore, the path information generation unit 208 can generate path information in a manner corresponding to the magnitude of the calculated path coefficient, and the display control unit 209 can thereby display path information in a manner corresponding to the magnitude of the calculated path coefficient.
  • the effect of allowing the user to easily recognize paths that have a high path coefficient and are likely to be useful can be obtained.
  • the path information generating unit 208 may also generate path information in a format according to the number of elements included in the detected path. This allows the display control unit 209 to display the path information corresponding to the detected path in a format according to the number of elements included in the path. For example, the path information generating unit 208 may set the thickness of the lines of the directed edges in the path information to a thickness proportional to the number of elements included in the path, or may vary the thickness of the lines of the directed edges depending on whether the number of elements is equal to or greater than a predetermined threshold.
  • the information processing device 2 in addition to the effect of the information processing device 1 according to the exemplary embodiment 1, the effect of allowing the user to easily recognize the difference in the number of elements contained in each path can be obtained.
  • the path information generating unit 208 may generate different types of path information depending on whether or not the path satisfies a predetermined condition. Whether or not the path satisfies the predetermined condition is determined by the condition satisfaction determining unit 206. This allows the display control unit 209 to display paths that satisfy the predetermined condition in a manner that makes them distinguishable from paths that do not satisfy the condition.
  • the effect of allowing the user to easily distinguish between paths that satisfy a predetermined condition and paths that do not satisfy the condition can be obtained.
  • the information processing device 2 also includes a cost calculation unit 207 that calculates the cost required to change a predetermined amount of the variable of the element at the end of each of the multiple paths detected by the path detection unit 203. Therefore, the path information generation unit 208 can generate path information in a manner corresponding to the magnitude of the calculated cost, and the display control unit 209 can display the path information in a manner corresponding to the calculated cost.
  • the effect of allowing the user to easily recognize the difference in the cost required for changing the variable of the element at the end point of the path by a specified amount for each path can be obtained.
  • FIG. 6 is a diagram showing another example of a display screen for path information.
  • the example screen b1 shown in Fig. 6 also includes an object b11 for accepting the designation of an element to be an end point, and an object b12 for accepting the designation of an element to be a start point, similar to the objects a1 and a2 in the example screen of Fig. 4.
  • the example screen b1 includes objects b13 to b15 in addition to the objects b11 and b12.
  • the path information in the example of Fig. 6 is each object and character string displayed in the portion surrounded by a dashed line in the example screen b1.
  • Object b13 is an object that indicates the specified starting element, and also indicates the starting element of the detected path.
  • Arrow objects b14 and b15 are displayed in association with object b13. These objects indicate the presence of elements connected downstream of the element of object b13, and are objects that accept input operations to display those elements.
  • object b14 is associated with a character string indicating that the index value calculated for the path corresponding to object b14 is large
  • object b15 is associated with a character string indicating that the index value calculated for the path corresponding to object b15 is medium.
  • the path information generation unit 208 can classify the magnitude of the index values calculated by the change index calculation unit 204 into large, medium, small, etc., and generate such path information based on the classification results.
  • object b21 indicates that the element connected downstream of the element "promotion" shown in object b13 in screen example b1 is "package".
  • object b22 is an object for displaying elements connected upstream of object b21. Therefore, when a user operation is performed to select object b22, the display state of screen example b1 is restored.
  • object b23 is an object for displaying elements connected downstream of object b21. Therefore, when a user operation is performed to select object b23, the dashed line portion in screen example b1 becomes the display state of b3.
  • b3 includes objects b31 and b32.
  • object b321 indicates that the element connected downstream of the element "Package" of object b21 shown in b2 is "Product Purchase.”
  • Object b31 also indicates that this element is the object specified as the end point.
  • Object b32 is an object for displaying the element connected upstream of object b31. Therefore, when a user operation is performed to select object b31, the display state returns to b2.
  • the dashed line portion in screen example b1 transitions to b4.
  • the display content of b4 is generally similar to b2, but differs in that object b41 in b4 indicates the element "price" and that a character string is displayed indicating that a path with a medium index value is being displayed.
  • the display content of b5 is generally similar to b3, but differs from b3 in that b5 displays a character string indicating that a path with a medium index value is being displayed.
  • the display control unit 209 does not necessarily need to display all elements of the detected path at once, and may display each element individually in response to a user operation, etc. Also, as in the example of FIG. 6, when displaying the components of each path, the display control unit 209 may vary the display mode of the path information by displaying a character string that indicates the characteristics of the path.
  • FIG. 7 is a diagram showing yet another example of a path information display screen.
  • the example screen shown in Fig. 7 also includes an object c1 for receiving the designation of an element to be an end point, and an object c2 for receiving the designation of an element to be a start point, similar to the objects a1 and a2 in the example screen of Fig. 4.
  • the example screen of Fig. 7 also includes path information c3.
  • Path information c3 shows paths using nodes and directed edges, similar to a causal graph. Specifically, path information c3 shows path 1, which goes from the element “nearby park” through the element “childcare support” to “livability”, and path 2, which goes from the element “nearby park” through the element “lots of greenery” to "livability”.
  • path information c3 the distance between nodes (which can also be said to be the length of a directed edge) is a distance that corresponds to the expected time it takes for a change in an upstream element to be reflected in a downstream element.
  • each detected path is displayed in a manner that corresponds to the propagation time of the change on each path, so it can be said that the shortest route shown in path information c3 indicates the path that is expected to produce the quickest effect. This allows the user to intuitively recognize which of multiple paths is expected to produce the earliest effect.
  • the distance from the element "nearby park” to the element “childcare support” in path 1 is shorter than the distance from the element "nearby park” to the element "lots of greenery” in path 2.
  • the above predicted time indicates the time cost of the path, and is calculated by the cost calculation unit 207. Therefore, the path information generation unit 208 can determine the distance between nodes based on the cost calculated by the cost calculation unit 207, i.e., the predicted time, and generate such path information. The path information generation unit 208 can also generate similar path information based on costs other than time.
  • FIG. 8 is a diagram showing yet another example of a path information display screen.
  • the example screen shown in Fig. 8 also includes an object d1 for accepting the designation of an element to be an end point, similar to the object a1 in the example screen of Fig. 4.
  • the example screen of Fig. 8 also includes objects d2 and d3, and path information d4.
  • Objects d2 and d3 are both objects for accepting the designation of the element to be used as the starting point, and are also objects that indicate the designated element. In other words, in the example of Figure 8, it is assumed that two elements have been designated as the starting point.
  • Path information d4 shows paths using nodes and directed edges, similar to a causal graph. Specifically, path information d4 shows path 1 from the element "cleaning state” through the element “good nearby parks” to “livability”, and path 2 from the element “clean river water” through the element “good nearby parks” to “livability”.
  • path information d4 nodes corresponding to each of the above elements are displayed on a coordinate plane with the vertical axis representing expected cost and the horizontal axis representing the expected time of effect. That is, in path information d4, the vertical position of each path indicates the cost expected to be required to improve the "livability" element at the end point by a specified amount through that path. In the example of Figure 8, it can be seen that the expected cost of path 2 is about twice that of path 1. Note that the cost in the "expected cost” in Figure 8 is a monetary cost, and is calculated by the cost calculation unit 207.
  • the horizontal distance between nodes indicates the expected time until a change in an element corresponding to an upstream node is reflected in an element corresponding to a downstream node. This expected time is also calculated by the cost calculation unit 207.
  • the horizontal distance from the element "The river water is clean” to the element "Nearby parks are well-equipped” is longer than the horizontal distance from the element "Cleanliness” to the element “Nearby parks are well-equipped.”
  • the path information generating unit 208 may set multiple evaluation axes for each path or its components, generate path information in which each path is displayed on the evaluation axes, and the display control unit 209 may display the path information. This allows the user to easily recognize multiple characteristics of each path.
  • the type of evaluation axis to be set is arbitrary and is not limited to the example of FIG. 8.
  • the path information generating unit 208 may set evaluation axes for the index values calculated by the change index calculation unit 204.
  • Fig. 9 is a flow diagram showing the flow of the analysis support method. Note that Fig. 9 shows processing in a state where the causal analysis by the causal analysis unit 202 is completed and causal relationships between multiple elements are identified.
  • the reception unit 201 receives the designation of at least one element from among the multiple elements that have been causally analyzed.
  • the display control unit 209 may display a list of the multiple elements that have been causally analyzed. In this case, the reception unit 201 receives the designation of the user's desired element from among the elements shown in the list.
  • the path detection unit 203 detects paths that include the element specified in S21 from among paths consisting of multiple elements that are causally linked. Then, in S23, the path detection unit 203 determines whether or not multiple paths have been detected. If the determination in S23 is YES, the process proceeds to S24.
  • the path information generating unit 208 generates path information indicating one path detected in S22.
  • the processes of S24 to S27 may also be performed if the result of S23 is YES. In this case, even if only one path is detected in S22, the path information generating unit 208 generates path information indicating that path in a manner according to the results of the processes of S24 to S27.
  • the path coefficient calculation unit 205 calculates a path coefficient for each of the multiple paths detected in S22. Note that the weight values used to calculate the path coefficients can be calculated using known causal inference or causal search methods.
  • the change index calculation unit 204 calculates, for each of the multiple paths detected in S22, the amount of change in the value of the element at the end of the path when the value of the element at the start of the path is changed, as an index value indicating the degree of change in the value of the element.
  • the path coefficient calculated in S24 is used to calculate the amount of change. Note that, as mentioned above, the index value indicating the degree of change is not limited to the amount of change, and may be the rate of change, etc.
  • condition satisfaction determination unit 206 determines whether each of the multiple paths detected in S22 satisfies a predetermined condition.
  • the cost calculation unit 207 calculates the cost required to change the value of the element at the end of each of the multiple paths detected in S22 by a specified amount. Note that the order in which the processes of S24 to S27 are performed is arbitrary, and some or all of these processes may be performed in parallel. Also, it is not necessary to perform all of the processes of S24 to S27, and only some of them may be performed.
  • the path information generating unit 208 generates path information indicating each path in a manner appropriate to each path, based on the results of the processes in S24 to S27. Note that the path information generating unit 208 does not necessarily need to use all of the results of the processes in S24 to S27. For example, the path information generating unit 208 may generate path information based on some or all of the results of S24 to S27, in accordance with a user specification. Also, for example, the path information generating unit 208 may generate multiple types of path information based on the results of the processes in S24 to S27.
  • the display control unit 209 causes the display unit 24 to display the path information generated in S28. This ends the processing in FIG. 9. Note that, as described above, multiple pieces of path information may be generated in S28, and when multiple pieces of path information are generated, the display control unit 209 may display the multiple pieces of path information. Furthermore, the display control unit 209 may switch the path information to be displayed in response to a user operation.
  • the execution entity of each process described in the above embodiment is arbitrary and is not limited to the above example.
  • the functions of the information processing devices 1 and 2 can be realized by multiple devices (which can also be called processors) that can communicate with each other.
  • processors which can also be called processors
  • each process described in the flowcharts of Figures 2 and 9 can be shared and executed by multiple processors.
  • the execution entity of the analysis support method in the above embodiment may be one processor or multiple processors.
  • Some or all of the functions of the information processing devices 1 and 2 may be realized by hardware such as an integrated circuit (IC chip), or may be realized by software.
  • the information processing devices 1 and 2 are realized, for example, by a computer that executes instructions of a program, which is software that realizes each function.
  • a computer that executes instructions of a program, which is software that realizes each function.
  • An example of such a computer (hereinafter referred to as computer C) is shown in Figure 10.
  • Computer C has at least one processor C1 and at least one memory C2.
  • Memory C2 stores a program (analysis support program) P for operating computer C as information processing device 1 or 2.
  • processor C1 reads and executes program P from memory C2, thereby realizing each function of information processing device 1 or 2.
  • the processor C1 may be, for example, a CPU (Central Processing Unit), GPU (Graphic Processing Unit), DSP (Digital Signal Processor), MPU (Micro Processing Unit), FPU (Floating point number Processing Unit), PPU (Physics Processing Unit), TPU (Tensor Processing Unit), quantum processor, microcontroller, or a combination of these.
  • the memory C2 may be, for example, a flash memory, HDD (Hard Disk Drive), SSD (Solid State Drive), or a combination of these.
  • Computer C may further include a RAM (Random Access Memory) for expanding program P during execution and for temporarily storing various data.
  • Computer C may further include a communications interface for sending and receiving data to and from other devices.
  • Computer C may further include an input/output interface for connecting input/output devices such as a keyboard, mouse, display, and printer.
  • the program P can also be recorded on a non-transitory, tangible recording medium M that can be read by the computer C.
  • a recording medium M can be, for example, a tape, a disk, a card, a semiconductor memory, or a programmable logic circuit.
  • the computer C can obtain the program P via such a recording medium M.
  • the program P can also be transmitted via a transmission medium.
  • a transmission medium can be, for example, a communications network or broadcast waves.
  • the computer C can also obtain the program P via such a transmission medium.
  • An information processing device comprising: a receiving means for receiving the designation of at least one element among a plurality of elements that have been causally analyzed; a path detection means for detecting a path that includes the designated element among paths consisting of a plurality of the elements that are causally connected; and a display control means for displaying path information indicating each path in a manner corresponding to each path when the path detection means detects a plurality of the paths that have a common end point element.
  • (Appendix 2) 2. The information processing device according to claim 1, wherein the display control means displays the path information on a causal graph in which a plurality of the elements are nodes and causal relationships between the elements are indicated by directed edges.
  • Appendix 3 An information processing device as described in Appendix 1 or 2, further comprising a change index calculation means for calculating, for each of the multiple paths detected by the path detection means, an index value indicating the degree of change in the value of the element at the end point of the path when the value of the element at the start point of the path is changed, and the display control means displays the path information in a manner corresponding to the index value indicating the degree of change.
  • Appendix 4 An information processing device as described in any of Appendix 1 to 3, further comprising a path coefficient calculation means for calculating a path coefficient for each of the multiple paths detected by the path detection means, and the display control means displays the path information in a manner corresponding to the magnitude of the path coefficient.
  • Appendix 7 An information processing device as described in any of Appendices 1 to 6, further comprising a cost calculation means for calculating a cost required to change a value of an element at the end point of each of the multiple paths detected by the path detection means by a predetermined amount, and the display control means displays the path information in a manner corresponding to the cost.
  • An analysis support method including: receiving, by at least one processor, the designation of at least one element among a plurality of elements that have been causally analyzed; detecting a path including the designated element among paths consisting of the plurality of elements that are causally connected; and, when a plurality of paths that share the element as an end point are detected, displaying path information indicating each path in a manner corresponding to each path (Appendix 9).
  • An analysis support program that causes a computer to function as: a receiving means for receiving the designation of at least one element among a plurality of elements that have been causally analyzed; a path detection means for detecting a path that includes the designated element among paths consisting of a plurality of the elements that are causally connected; and a display control means for displaying path information indicating each path in a manner corresponding to each path when the path detection means detects a plurality of paths that have a common end point element.
  • An information processing device including at least one processor, the processor executing a process of accepting designation of at least one element among a plurality of elements that have been causally analyzed, a process of detecting a path including the designated element among paths made up of the plurality of elements that are causally connected, and a process of displaying path information indicating each path in a manner corresponding to each path when a plurality of paths that share a common end point element are detected.
  • the information processing device may further include a memory, and the memory may store a program for causing the processor to execute the process of accepting element designation, the process of detecting a path, and the process of displaying path information.
  • the program may also be recorded on a computer-readable, non-transitory, tangible recording medium.
  • Information processing device 11 Reception unit (reception means) 12 Path detection unit (path detection means) 13 Display control unit (display control means) 2 Information processing device 201 Reception unit (reception means) 203 Path detection unit (path detection means) 204 Change index calculation unit (change index calculation means) 205 Path coefficient calculation unit (path coefficient calculation means) 207 Cost calculation unit (cost calculation means) 209 Display control unit (display control means)

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • Pure & Applied Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • Algebra (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Computational Mathematics (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Probability & Statistics with Applications (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
PCT/JP2022/047046 2022-12-21 2022-12-21 情報処理装置、分析支援方法、および分析支援プログラム Ceased WO2024134795A1 (ja)

Priority Applications (2)

Application Number Priority Date Filing Date Title
JP2024565468A JPWO2024134795A1 (https=) 2022-12-21 2022-12-21
PCT/JP2022/047046 WO2024134795A1 (ja) 2022-12-21 2022-12-21 情報処理装置、分析支援方法、および分析支援プログラム

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/JP2022/047046 WO2024134795A1 (ja) 2022-12-21 2022-12-21 情報処理装置、分析支援方法、および分析支援プログラム

Publications (1)

Publication Number Publication Date
WO2024134795A1 true WO2024134795A1 (ja) 2024-06-27

Family

ID=91588185

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2022/047046 Ceased WO2024134795A1 (ja) 2022-12-21 2022-12-21 情報処理装置、分析支援方法、および分析支援プログラム

Country Status (2)

Country Link
JP (1) JPWO2024134795A1 (https=)
WO (1) WO2024134795A1 (https=)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2008226268A (ja) * 2008-05-09 2008-09-25 National Institute Of Advanced Industrial & Technology モデル作成装置、情報分析装置、モデル作成方法、情報分析方法、およびプログラム
WO2014184934A1 (ja) * 2013-05-16 2014-11-20 株式会社日立製作所 障害分析方法、障害分析システム及び記憶媒体
JP2021002335A (ja) * 2019-06-21 2021-01-07 パロ アルト リサーチ センター インコーポレイテッド 高次元センサデータにおける異常事象の自動化された根本原因分析を実行する方法及びシステム
JP2021111141A (ja) * 2020-01-10 2021-08-02 国立大学法人京都大学 特徴ネットワーク抽出装置、コンピュータプログラム、特徴ネットワーク抽出方法及びベイジアンネットワーク分析方法
WO2022254802A1 (ja) * 2021-05-31 2022-12-08 日立Astemo株式会社 改善システム、改善方法

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2008226268A (ja) * 2008-05-09 2008-09-25 National Institute Of Advanced Industrial & Technology モデル作成装置、情報分析装置、モデル作成方法、情報分析方法、およびプログラム
WO2014184934A1 (ja) * 2013-05-16 2014-11-20 株式会社日立製作所 障害分析方法、障害分析システム及び記憶媒体
JP2021002335A (ja) * 2019-06-21 2021-01-07 パロ アルト リサーチ センター インコーポレイテッド 高次元センサデータにおける異常事象の自動化された根本原因分析を実行する方法及びシステム
JP2021111141A (ja) * 2020-01-10 2021-08-02 国立大学法人京都大学 特徴ネットワーク抽出装置、コンピュータプログラム、特徴ネットワーク抽出方法及びベイジアンネットワーク分析方法
WO2022254802A1 (ja) * 2021-05-31 2022-12-08 日立Astemo株式会社 改善システム、改善方法

Also Published As

Publication number Publication date
JPWO2024134795A1 (https=) 2024-06-27

Similar Documents

Publication Publication Date Title
US10599449B1 (en) Predictive action modeling to streamline user interface
JP4678943B2 (ja) 新製品の初期投入量予測方法、新製品の初期投入量予測装置及び記録媒体
CN108337316A (zh) 信息推送方法、装置、计算机设备及存储介质
TW202008237A (zh) 針對新場景的預測模型訓練方法及裝置
WO2019114423A1 (zh) 对模型预测值进行融合的方法、装置和设备
JP4465417B2 (ja) 顧客セグメント推定装置
CN115344663B (zh) 一种业务逻辑的知识表示和推演方法及装置
CN113469482A (zh) 用于数据分析的方法、设备以及计算机可读存储介质
JP2013532874A (ja) 履歴データに基づく適合性の確度の決定
EP3279806A1 (en) Data processing method and apparatus
JP7407543B2 (ja) 情報処理方法、プログラム、及び情報処理装置
US10537801B2 (en) System and method for decision making in strategic environments
Jang et al. Detection and prediction of house price bubbles: Evidence from a new city
CN110245999A (zh) 信息推荐方法、信息显示方法、装置及计算设备
JP6159056B2 (ja) 選択プログラム、選択方法及び選択装置
Araman et al. Crowdvoting the timing of new product introduction
CN115797005A (zh) 预测模型的训练、商品推荐方法、装置、电子设备及介质
CN111026973A (zh) 一种商品兴趣度预测方法、装置及电子设备
JP2009251779A (ja) 販売数予測システム、販売数予測方法および販売数予測プログラム
KR102156521B1 (ko) 의류 쇼핑몰 전자상거래 시스템
WO2024134795A1 (ja) 情報処理装置、分析支援方法、および分析支援プログラム
US20120136822A1 (en) Key performance indicator feedback in rule-based system
WO2024134797A1 (ja) 情報処理装置、シミュレーション方法、およびシミュレーションプログラム
CN106682888B (zh) 业务信息的处理方法和处理装置
CN112381617A (zh) 网络购物方法及装置

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: 22969183

Country of ref document: EP

Kind code of ref document: A1

WWE Wipo information: entry into national phase

Ref document number: 2024565468

Country of ref document: JP

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 22969183

Country of ref document: EP

Kind code of ref document: A1