JP2006260333A - Flow retrieval method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To automatically form a retrieval table, in a flow retrieval method for performing retrieval of a flow showing a processing flow, based on a flow showing a processing flow designated by a user, and retrieve a DB based on the retrieval table, and present flows in descending order of similarity, depending on the action parts present in the flow, attributes and sub-attributes thereof, or the like. <P>SOLUTION: This method comprises steps of extracting and setting, based on the flow showing a processing flow designated by the user, action parts constituting this flow and attributes thereof as a retrieval table; retrieving, based on the retrieval table, the DB in which action parts constituting a flow for each formed flow and attributes thereof are extracted and registered to extract flows having high similarity; and presenting the retrieved flows with high similarity. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a flow search method for searching for a flow representing a processing flow.

  Conventionally, in a workflow system, a series of application routes is realized by connecting each route component, and common route components can be used for various application scenes.

  At this time, when searching for a route component to be designed from a variety of already designed route component groups, a list matching the search condition is displayed in the text search.

Further, there is a technique in which keywords are registered in advance using a workflow as a template, a list of templates that match a search condition input by a user is displayed, and a template is selected (Patent Document 1).
JP 2001-125984 A

  In the former described above, a text search is used to select a part, and the contents of a part (for example, where a conditional branch enters and what action is taken) are detailed description and design of the part. There was a problem that a lot of time and effort was required to refer to and confirm the book separately.

  In the latter case, text search can only be performed for the content registered in advance in the template, and the detailed content of the component requires a lot of trouble to confirm separately by referring to the detailed description of the component or a design document. There was a problem of becoming.

  In order to solve these problems, the present invention creates a search table in which action parts, attributes, sub-attributes, etc. are extracted based on a search flow (workflow or the like) input from the user, and The DB is searched based on the table and the flows are presented in descending order of similarity.

  The present invention automatically creates a search table based on a flow (workflow or the like) representing a processing flow designated by a user, searches a DB based on the search table, and an action that exists in the flow It is possible to present a flow in descending order of similarity depending on parts, attributes, and sub-attributes.

  The present invention creates a search table in which action parts, attributes, sub-attributes, etc. are extracted based on a search flow (workflow or the like) input by a user as described below. In addition, the DB is searched and the flows are presented in descending order of similarity.

FIG. 1 shows a system configuration diagram of the present invention.
In FIG. 1, a processing apparatus 1 is connected to a network 12 to create and search a search table in response to a flow search request from a large number of terminals 13, and returns a search result. In this example, the search condition receiving means 2, the similarity search means 3, the search result display means 4, the search sub-table 5, the search condition management table 6, the root component configuration sub-table management DB 7, The action component management DB 8, the action component image management DB 9, the sub-table management DB 10, the similarity search work table 11, and the like are included.

  The search condition accepting unit 2 accepts a flow search condition via the network from the terminal 13 (see FIGS. 2 to 5).

  The similarity search means 3 performs a similarity search based on the search sub-table 5 generated from the flow (see FIGS. 6 to 9).

  The search result display means 4 displays the search result (see FIG. 10 and FIG. 4B).

  The search sub-table 5 is a search table generated based on a flow for which a search request has been made (see FIG. 3).

  The search condition management table 6 sets and manages search conditions obtained by analyzing a flow for which a search request has been made (see (c) of FIG. 3).

  The root part configuration subtable management DB 7 registers and manages the subtables constituting the root part (see FIG. 5A).

  The action component management DB 8 registers and manages action component attributes (see FIG. 5B).

  The action component image management DB 9 registers and manages action component images (images displayed on the flow) (see FIG. 5C).

  The sub-table management DB 10 sets and manages action component types and attributes and further sub-attributes in association with sequence numbers (see (b) to (e) of FIG. 7).

  The similarity search work table 11 is a work table for calculating by adding a score for each sub-table in association with a root part at the time of similarity search (similar search work table 1 to which the score is added and managed, and a sequence number) (See FIG. 8 and FIG. 9).

  The network 12 is a network through which a large number of terminals 13 and the processing apparatus 1 transmit and receive data such as search requests and result responses, and is, for example, the Internet or a LAN.

  The terminal 13 is operated by the user, creates a flow on the screen, requests a search (similar search) from the processing device 1 via the network 12, receives a search result, and displays it on the screen. For example, a personal computer.

Next, the overall configuration of the configuration of FIG. 1 will be described in the order of the flowchart of FIG.
FIG. 2 shows a flowchart (overall) for explaining the operation of the present invention.

  In FIG. 2, S1 performs search condition reception processing. This is because the user operates the terminal 13 shown in FIG. 1 to select a copy from the list of action parts on the left side with a mouse on the screen shown in FIG. After creating the flow as shown in the figure by repeating the input, the processing device 1 is configured with the flow data via the network 12 in response to pressing the similar route part search button in the lower right. The search condition accepting means 2 to accept

  In S2, a similarity search process is performed. This is the processing device 1 received in S1, and the similarity search means 3 performs the search sub-table 5 and the search sub-table 5 and the flow for which the search request is made (for example, the flow of FIG. 4A) according to the flowchart of FIG. After the search condition management table 6 is created, a similarity search is performed according to the flowchart of FIG.

  In S3, search result display processing is performed. This is because the result of the similarity search in S2 is displayed on the screen of the terminal 13, for example, as shown in the screen of FIG. 4B described later, registered based on the result of the search in descending order of similarity. The generated flow (similar flow) was displayed.

  As described above, when the user creates the flow shown on the right side of the screen of FIG. 4A on the screen of the terminal 13 and performs a similar search, the processing device 1 also displays the flow for which the search request has been made. In addition, the search sub-table 5 and the search condition management table 6 are automatically created, the registered tables are searched based on the automatically created tables, the score is calculated, and the search results are ranked in descending order of the score. This flow can be displayed on the screen of the terminal 13 as shown in FIG. Details will be sequentially described below.

  FIG. 3 is a flowchart for explaining the operation of the present invention (creation of search sub-table). This is based on the search target flow (for example, the flow on the right side of FIG. 4A) received in the search condition reception process of FIG. 2 described above, and the search sub-table 5 and the search condition management table 6. 6 is a flowchart describing in detail the procedure for automatically creating a file.

  3, S11 acquires the positional relationship of the action component objects such as Δ, ○, and mouth arranged on the screen, starting from the upper left side of the screen (sequence No. 1). For example, for the flow on the right side of FIG. 4A in which a search request is made by the user, “▽” is acquired here as the starting point on the upper left side.

  In S12, the number (L) of action part objects on the right side of the screen (flow end) is acquired and set in the “Route Number” column of the search condition management table. For example, the number of routes from the starting point “▽” on the right screen of FIG. 4A to the two action part objects “mouth” on the right end, in this case, “2” is acquired. 3 is set to “2” in the route number column of the search condition management table 6 in (c), and the fact that the number of routes from the starting point of the flow to the end action component is two is stored.

  S13 secures and initializes an area of L sub-tables 5 for search. This is because the number of routes L acquired in S12, here L = 2, the two search subtails 5 acquire areas as shown in (b-1) and (b-2) of FIG. (Here, since the number from the starting point to the action component at the end of the right end is five on the screen on the right side of FIG. 4A, (b-1) and (b− As shown in 2), an area provided with five records of sequence Nos. 1, 2, 3, 4, and 5 is acquired and initialized).

In S14, a variable n = 1 is initially set.
In S15, the number (M) of action component objects from the starting point to the terminal n is acquired and set in the “route n action number” of the search condition management table. Since this is the first time when n = 1, the number of actions from the leftmost starting point “▽” on the right screen of FIG. 4A to the rightmost action part object “mouth” on the upper route = 5 is acquired and set as the number of route 1 actions = 5 in the search condition management table 6 of FIG.

In S16, a variable k = 1 is initially set.
S17 searches the action component image management DB 9 with the image of the kth action component object, and sets the corresponding action component in the kth “action component” column of the search subtable 5 (n). Since this is the first time when k = 1, the action part image of FIG. 5C is used for the image “▽” of the k = 1st action part object on the screen on the right side of FIG. The management DB 9 is searched, and the action component “draft” is obtained here, so that the “draft” is assigned the sequence number of (b-1) in FIG. 1 is set in the column of action parts as shown in the figure. Similarly, when k = 2, “branch (by amount)”, when k = 3 “approved (affiliation)”, when k = 4 “approved (general affairs)”, when k = 5 “ “Reserve acquisition” is set as shown in (b-1) of FIG.

  In step S18, it is determined whether an attribute is specified for the kth action component object. For example, it is determined whether or not the attribute is specified for the starting point “▽” on the screen on the right side of FIG. 4A which is the k = 1st action component object. Here, since attribute 3 (person in charge) is set for the starting action part object “▽”, the answer is YES, and the attribute specified in S19 is the k-th “in the sub-table for search (n)”. Set in the “Attribute” column. Here, the set attribute 3 (person in charge) of the action component object “「 ”as the starting point is set to the sequence No. of (b-1) in FIG. As shown in the figure, “person in charge” is set in the attribute 3 of the record of the action component “draft” of 1 origin. Here, the “person in charge” in “▽” of the action part object at the starting point in FIG. 4A refers to the action part management DB 8 in FIG. 5B and is found to belong to “attribute 3”. Therefore, “attribute 3” is set in the attribute field of (b-1) in FIG. Similarly, when k = 2, the sequence number of (b-1) in FIG. “Attribute 3” is set in the attribute field of the action component “Branch (by amount)” in the record 2. Then, the process proceeds to S20. On the other hand, if NO in S18, the process proceeds to S20.

  In S20, it is determined whether or not a sub attribute is designated for the kth action component object. For example, it is determined whether or not a sub-attribute is specified for the starting point “▽” on the screen on the right side of FIG. 4A which is k = 1st action component object. Here, since the sub-attribute (below) is set for the starting action part object “▽”, the answer is YES, and the sub-attribute specified in S21 is the k-th “in the search sub-table (n)”. Set in the “Sub attribute” field. Here, the set sub-attribute (below) of the starting action part object “」 ”is set to the sequence number of (b-1) of FIG. As shown in the figure, “below” is set in the sub-attribute of the record of the action component “draft” of 1 origin. Here, “below” in “▽” of the action part object of the starting point in FIG. 4A refers to the action part management DB 8 in FIG. 5B and is found to belong to “sub attribute”. Therefore, “below” is set in the sub-attribute column of (b-1) in FIG. Similarly, when k = 2, the sequence number of (b-1) in FIG. In the second record, “above” is set in the sub attribute column of the action component “Branch (by amount)”. Then, the process proceeds to S22. On the other hand, if NO in S20, the process proceeds to S22.

In S22, k = k + 1.
S23 determines whether k> M. This is because a search sub-table when k exceeds M (the number from the starting point to the end action part object, here M = 5) and there are attributes and sub-attributes up to the end action part object. 5 determines whether the setting has been completed. If YES, the process proceeds to S24. In the case of NO, the process from S17 onward is repeated, and the attributes and sub-attributes of the next action part are extracted and set in the search sub-table 5 if they are in the flow.

In S24, n = n + 1.
In S25, it is determined whether n> L. This determines whether n has exceeded L (the number L of the action component objects at the end of the flow) and the setting to the search sub-table 5 has been completed up to the end action component object. If YES, the process ends. In the case of NO, the steps from S15 are repeated, and the next action component is extracted from the flow and set in the search sub-table 5 repeatedly.

  By the above procedure, for example, the flow on the screen on the right side of FIG. 4A can be used as an input, and the search sub-table 5 of FIGS. 3B-1 and 3B-2 can be automatically created. It becomes.

FIG. 4 shows an example of a screen (flow) according to the present invention.
FIG. 4A shows an example of an input screen. This is a screen example for the user to specify copy + attribute (sub-attribute) from the left action component list and create a flow shown on the right screen. In order to create, from the action parts list, first create ▽ (draft) by selecting it with the mouse and dragging it to the position shown on the screen on the right side. An external attribute menu is displayed and selected or input, and here, the state after selecting and specifying the attribute “person in charge” and the sub attribute “below” is shown. Similarly, by copying the action part object “Branch (by amount)” and specifying the attributes “4999”, sub-attributes “above” and “below” as shown in the figure, the user can execute the flow shown in the figure. Create

  As described above, a flow is created by specifying copy + attribute (sub-attribute) from the action component list, and the search sub-table 5 is automatically created based on the flow, and the similarity is high from the DB accumulated in the past. It is possible to request a similarity search for flows.

  FIG. 4B shows an example of a search result screen. This is because the search sub-table 5 is automatically created in the flowchart of FIG. 3 based on the flow on the right screen of FIG. 4A, and is created in advance based on the created search sub-table 5 The DBs are searched and the flows are displayed in descending order of similarity (the upper flow having the highest similarity and the lower flow having the next highest similarity).

  As described above, since flows are displayed in descending order of similarity, it is possible to select and divert the flow created in (a) of FIG. Become.

FIG. 5 shows an example DB of the present invention.
FIG. 5A shows an example of a root part configuration subtable management DB. The root part configuration subtable management DB 7 shown in the figure registers and manages the subtables constituting the root part, and registers and manages the following information shown in association with each other.

・ Root part number:
・ Root part name:
Sub-table 1:
Sub-table 2:
...
Here, the root part number is a unique number assigned to the root part. The root part name is a name given to the root part. Sub-tables 1, 2,... Are registered with all the sub-tables 1, 2,. Sub-tables 1, 2,... Are the numbers of all sub-tables constituting the route part. For example, the root part number “R01” is composed of two parts, sub-table 1 and sub-table 2. .

  FIG. 5B shows an example of an action component management DB. The action part management DB 8 shown in the figure is registered with attributes 1, 2, 3, 4, 5... Of the action parts. Here, the following information shown in FIG. is there.

・ Action parts:
・ Attribute 1:
・ Attribute 2:
・ Attribute 3:
・ Attribute 4:
・ Attribute 5:
...
Here, the action component defines contents, ranges, and the like representing the attributes 1, 2, 3, 4, 5,... Of the action component on the flow. The attribute “temporary employee” is defined as attribute 1, and the attribute “general employee” is defined as attribute 2. For example, the “person in charge” of the action part “draft” of the flow on the screen on the right side of FIG. 4A described above is the action part “draft” of the action part management DB 8 of FIG. 5B. Corresponds to the “section chief” in the entry of “No. 3” and becomes “attribute 3”, so that the sequence number “b-1” in FIG. “Attribute 3” is set in the attribute column of the action part “draft” of the first entry.

  FIG. 5C shows an example of an action part image management DB. The action component image management DB 9 shown in FIG. 4 refers to the action component image management DB 9 in FIG. 5C for the “▽” in the flow on the right side of the screen of FIG. “▽” is determined to be the action component “draft”, and the sequence number of the search sub-table 5 in FIG. The “draft” is set in the action part column of one entry.

  Next, in accordance with the order of the flowchart of FIG. 6, (b-1) of FIG. 3 created from the flow of FIG. 4A according to the flowchart of FIG. ) And (b-2) based on the two search sub-tables 5 and the search condition management table 6, the procedure for performing a similar search from the DB (FIGS. 5 and 7) will be described in detail below.

  FIG. 6 is a flowchart for explaining the operation of the present invention (similarity search). Here, ○ 1 (representing the number 1 with ○, the same applies hereinafter) to ○ 13 correspond to the specific examples shown by ○ 1 to ○ 13 in FIGS. 8 and 9, respectively.

  In FIG. 6, S31 reads the search condition management table 6, acquires the number of routes (L), and initializes the similar search work table 1 (see FIG. 6B) accordingly. Here, the number of routes (L = 2) of the search condition management table 6 in FIG. 3C described above is acquired, and the area of the similar search work table 1 in FIG. 6B is secured and initialized. (Refer to ○ 1 in FIG. 8).

  In S32, the root part configuration sub-table management DB 7 is searched, and “root part No.” number defining the same number of sub-tables as the number of roots is set to J, and each “root part No.” is set to the similarity search table 1. In the “root part No.” field, the registered root part configuration subtable management DB 7 in FIG. 5A described above is searched, and the number of roots (here, L = obtained in S31). The number of “root parts No” defining the same sub-table as in 2) is J (here, J = 2), and the root part No. with two sub-tables 1 and 2 is “R01”. “R02” and “R04” are set in the root part number column of the similarity search work table 1 in FIG. 6B as shown in FIG. In the flow of (a) of FIG. Three route part numbers having the same route number 2 as the extracted route number L = 2 (“R01”, “R02”, and “R04”) are extracted and set in the root part number column of the similarity search work table 1 (See circle 2 in FIG. 8).

In S33, a variable n = 1 is initially set.
In S34, a variable m = 1 is initially set.

      S35 reads the root part configuration subtable management DB 7 at the nth “root part No” in the similarity search work table 1, reads the mth subtable (subtable m) of the record, and the corresponding subtable. The final sequence number in the management DB 10 is j. The similar search work table 2 is initialized.

In S36, a variable k = 1 is initialized.
In S37, it is determined whether or not the action component is the same as the k-th action component in the search sub-table n. For example, it is determined whether or not the action component is the same as the k = 1st action component “draft” of the search sub-table n = 1 in FIG. 3B-1. If YES, the process proceeds to S38. If NO, the process proceeds to S44.

            In S38, “sequence No.” of the same action part in the root part configuration subtable management DB 7 is acquired and added to the similarity search work table 2. This is YES in S37, for example, the sequence number of the search sub-table 1 in (b-1) of FIG. A diagram generated for the upper route of the flow of FIG. 7A (to be described later) having the same action component as the action component “draft” of one entry (generated in the same manner as the flowchart of FIG. 3). 7 (b), the action part “draft” sequence No., here, “1” indicated by (a) is obtained, and the similarity search operation of FIG. 6 (c) is obtained. It is set to the first entry of sequence No. in table 2 (see circle 5 in FIG. 8).

            In S39, 1 is added to the n-th sub-table m similarity score of the similarity search work table 1. As indicated by ○ 6 in FIG. 8, 1 is added to the number of similarities in the sub-table 1 (R0101) of the root component (R01) that was the same action component “draft” in S38, and the action component “ Reflects the degree of similarity due to the “voting” matching.

            S40 determines whether the attributes of the same action component are the same. It is determined whether or not the attributes of the same action part “draft” added in S39 are the same. Here, the attribute “attribute 3” of the action part “draft” in FIG. It is determined whether or not the attribute “attribute 2” of the action part “draft” in (b) is the same. Here, since it becomes NO, it progresses to S42. On the other hand, in the case of YES, 1 is added to the number of similar points in the n-th sub-table m of the similarity search work table 1 in S41 (see ○ 6 in FIG. 8). Then, the process proceeds to S44.

              In S42, it is determined that the action parts are the same and the attributes are different in NO in S40, and further, if the attributes are different but the sub-attributes are considered, it is determined whether they are satisfied. Here, the attribute “attribute 3” of the action component “draft” in FIG. 3B-1 and the sub-attribute “below” are combined to become “below the chief” with reference to FIG. On the other hand, the attribute “attribute 2” and sub-attribute “none” of the action component “draft” in FIG. 7B become “general employee”, and the former “person in charge” is the latter “general employee”. Since it satisfies, it becomes YES, and 0.5 is added to the nth sub-table m similarity score of the similarity search work table 1 in S43 (0.5, which is a half when the action parts and attributes are matched) (see FIG. (See o10 of 8). Then, the process proceeds to S44.

In S44, k = k + 1.
In S45, it is determined whether k> j (final sequence number). If YES, the process proceeds to S46. On the other hand, in the case of NO, S37 and subsequent steps are repeated for the next sequence No.

      In S46, it is determined whether the sequence No. in the similarity search work table 2 is in ascending order. In the case of YES, the number of records in the similar search work table 2 is added to the number of similar points in the n-th sub-table m of the similar search work table 1 in S47 (see ○ 12 in FIG. 8, here, 5 points are added) To do). Then, the process proceeds to S48. On the other hand, if NO in S46, the process proceeds to S48.

In S48, m = m + 1 is set.
In S49, it is determined whether m> L (the number of routes, here 2). In the case of YES, since it is repeated by the number of routes L, the process is terminated, and the process proceeds to S50. On the other hand, in the case of NO, S35 and subsequent steps are repeated for the next route.

  In S50, the similarities of the n-th sub-tables 1 to (m-1) of the similarity search work table 1 are totaled and set to the total similarity (see circle 13 in FIG. 8).

In S51, n = n + 1.
In S52, it is determined whether n> j (the number of root parts No., here, 3). If YES, the process ends. In the case of NO, S34 and subsequent steps are repeated for the next route number No.

  As described above, the search sub-table in FIG. 3B and the search condition management table in FIG. 3C are automatically created based on the flow specified by the user (FIG. 4A). (FIG. 3), based on these tables, extract all root parts with the same number of routes (here, the number of routes = 2) from the root part configuration sub-table management DB 7 of FIG. Set in the work table 1, and if the action parts and attributes match for each root part set in the similarity search work table 1, 1 point is obtained. If the attributes are different but sub attributes are included, 0.5 points are satisfied. Further, if the appearance order is ascending order, the total number of similar points (here, 5 points) is calculated as the similarity, and displayed in the order of the root parts having the highest similarity (see, for example, FIG. 4B). ).

FIG. 7 shows a DB example of the present invention.
7A shows an example of a flow (route parts (R01 employee (day trip)), where the leftmost object part “draft” (general employee) is the starting point and the upper and lower routes. There are two routes.

  FIG. 7B shows an example of the sub-table management DB (R0101). The illustrated sub-table management DB (R0101) represents the upper route of the flow in FIG. For example, the top sequence No. For the action part type “draft” 1 and the attribute “attribute 2”, the first “▽” at the left end of the flow in FIG. 7A refers to the action part image management DB 9 in FIG. Corresponding to the action part “draft”, the “general employee” described inside ▽ refers to the action part management DB 8 in FIG. 5B, and the “general employee” of the action part “draft” is “ Since attribute 2 ″ is set, these are set. Similarly, when the second to fifth action parts, attributes, and sub-attributes are set, the result is as shown (the same as the upper flow in FIG. 4B).

  FIG. 7C shows an example of the sub-table management DB (R0102). The sub-table management DB (R0102) shown in the figure represents the route at the lower stage of the flow in FIG.

  FIGS. 7D and 7E show other examples of sub-table management DBs, respectively. The flow is shown in the lower part of FIG.

8 and 9 show specific examples in the flowchart of FIG. ◯ 1 to ◯ 13 in the figure indicate ◯ 1 to ◯ 13 in FIG. 6, respectively. 8 and 9 are described according to the order of the flow chart of FIG. For example, in FIG.
The leading circle 1 indicates an example of the similarity search work table 1 that has been initialized by securing the area in S31 of FIG.

    The second ○ 2 is the route part of the similar search target having the number of routes that matches the number of routes = 2 in the search condition management table 6 of FIG. 3C in S32 of FIG. The root part configuration subtable management DB 7 in (a) is retrieved, set, and set (here, the three root parts “R01” and “R02” shown in FIG. “R04” is set).

    The third circle 3 is created by creating a similarity search work table 2 having a sequence No. 5 in S35 of FIG.

    The fourth ○ 5 is the route component (R0101) to be searched that is the same as the action component “draft” of sequence No = 1 in FIG. 3B-1 in S38 of FIG. 6 (FIG. 7). The sequence number (here, “1”) of the action part “draft” (see (b)) is set in the similarity search work table 2.

    The fifth circle 6 is incremented by 1 because the action parts are the same in S39 of FIG.

    The sixth ◯ 10 is 0.5 in S43 of FIG. 6 because the attribute is different but is satisfied when sub-attributes are considered.

    ・ Seventh and eighth ○ 5 and ○ 6 set the sequence number = 2 of the second action part “branch (by amount)” and the action part “branch (by money)” matched. One point is added to 2.5.

    The ninth circle 12 is repeated for all actions (here, 5) existing in one route, the similarity is 9.5, and the appearance order of action parts is the same in S47 of FIG. 6 ( Since the similarity search work table 2) is in ascending order, 5 points are further added to the 9.5 to obtain 14.5.

  As described above, one route is calculated as a similar point (14.5), and is set in the similarity search table 1 as indicated by ◯ 12.

    The tenth circle 13 is obtained by calculating the similarities for each sub-table by repeating the circles 3 to 12 (third to ninth). Then, the sum of similarities = 26.5, which is the sum of the similarities in the subtables 1 and 2 of the route part No. “R01”, is calculated and set as shown in the figure.

    The eleventh (bottom row) repeats from ○ 5 (fourth) to ○ 12 (9th) for the route component (R02), and similarly, similarities are found for the sub-table 1 of the route component (R02). calculate.

    The first ○ 6 in FIG. 9 is repeated for the route component (R02) from ○ 5 (fourth) to ○ 12 (9th) in FIG. 8, and in the same manner, the subtable of the route component (R02). Similarities are calculated for 2, and here, the action component “draft” is the same and is incremented by one.

    In the second circle 8 in FIG. 9, the attribute “person in charge” is the same for the route component (R02) and is incremented by one.

・ In the third circle 6 in FIG. 9, the action part “branch (by amount)” is the same and +1 is added.
・ The fourth circle in FIG. 9 is different from “49999” in the attribute “9999”, but when looking at the sub-attribute, “9999”, “more than” corresponds to “4999”, “above”. Add 0.5 and set a total of 3.5.

    The fifth part of FIG. 9 calculates the similar sum of the root parts (R02) and the similarities of the root parts (R04) according to the above flow, and finally becomes as shown in the table of FIG. Of these, when the flow of the route component (R01) and the route component (R02) is displayed as a result in descending order of the similarity, the result is as shown in FIG.

FIG. 10 shows a flowchart (display) for explaining the operation of the present invention.
In FIG. 10, S 61 displays the search results in descending order of the “similarity total” score in the similarity search work table 1. This is displayed as a search result in the descending order of the similarity score of the fifth final similarity search work table 1 of FIG. 9 described above, for example, as shown in the screen of FIG. 4B described above. To do. In FIG. 4B, the flow of the upper first route component (R01) and the second route component (R02) is displayed. Then, from the displayed flow of FIG. 4B, the radio button of the flow to be diverted is selected, the diversion route component determination button is pressed, and an appropriate one is selected from the list of flows that are similar-searched based on the flow. The route component can be diverted.

(Appendix 1)
In a flow search method for searching for a flow representing a processing flow,
Based on the flow representing the processing flow specified by the user, extracting the action components and their attributes that constitute the flow and setting them in the search table;
Searching the registered DB by extracting the action parts and their attributes constituting the flow with respect to the already created flow based on the search table, and extracting a flow having a high degree of similarity;
Presenting the searched flow with high similarity.

(Appendix 2)
The number of routes to the end component starting from the first component constituting the flow is calculated, and the action component and its attribute are extracted for each route and set in the search table, respectively. 1. The flow search method according to 1.

(Appendix 3)
The flow search method according to appendix 1 or appendix 2, wherein in addition to the attribute of the action part, when the sub attribute is further set, the sub attribute is set together in the search table.

(Appendix 4)
The similarity is determined when the attribute matches the action component that constitutes the specified flow, and when the attribute does not match but the sub-attribute is specified and the sub-attribute is included, 4. The flow search method according to any one of appendix 1 to appendix 3, wherein a total score obtained by adding a predetermined score is used.

(Appendix 5)
In a flow search device that searches for a flow that represents a flow of processing,
Based on the flow representing the processing flow specified by the user, means for extracting the action parts and their attributes constituting the flow and setting them in the search table;
Based on the search table, for a flow that has already been created, an action component that constitutes the flow and a DB that has been registered by extracting the attribute, and a flow that extracts a flow having a high degree of similarity;
A flow search apparatus including means for presenting the searched flow having a high degree of similarity.

(Appendix 6)
The number of routes from the first component constituting the flow to the last component is calculated, and the action component and its attribute are extracted for each route and set in the search table. Item 6. The flow search device according to Item 5.

  The present invention automatically creates a search table based on a flow (workflow or the like) representing a processing flow designated by a user, searches a DB based on the search table, and an action that exists in the flow It is possible to present a flow in descending order of similarity depending on parts, attributes, and sub-attributes.

It is a system configuration diagram of the present invention. It is an operation | movement description flowchart (whole) of this invention. It is an operation | movement description flowchart (creation of the subtable for search) of this invention. It is an example of a screen (flow) of the present invention. It is DB example of this invention. It is operation | movement description flowchart (similarity search) of this invention. It is DB example of this invention. It is similar search explanatory drawing (the 1) of this invention. It is similar search explanatory drawing (the 2) of this invention. It is operation | movement description flowchart (display) of this invention.

Explanation of symbols

1: Processing device 2: Search condition receiving means 3: Similarity search means 4: Search result display means 5: Search sub-table 6: Search condition management table 7: Root component configuration sub-table management DB
8: Action parts management DB
9: Action parts image management DB
10: Sub-table management DB
11: Similarity search work table 12: Network 13: Terminal

Claims (3)

In a flow search method for searching for a flow representing a processing flow,
Based on the flow representing the processing flow specified by the user, extracting the action components and their attributes that constitute the flow and setting them in the search table;
Searching the registered DB by extracting the action parts and their attributes constituting the flow with respect to the already created flow based on the search table, and extracting a flow having a high degree of similarity;
Presenting the searched flow with high similarity.   The number of routes from the first component constituting the flow to the last component is calculated, and the action component and its attribute are extracted for each route and set in the search table. Item 2. The flow search method according to Item 1.   3. The flow search method according to claim 1, wherein when a sub attribute is set in addition to the attribute in the action part, the sub attribute is set together in the search table together. .

Images