Information processing method and device for child consultation centers, etc.

Abstract

To assist the personnel of a child consultation center, etc. in making decisions.SOLUTION: The present information processing method includes: a step (A) in which, when data having been inputted regarding at least one of a plurality of items including an item about information pertaining to a child abuse notice and an item about risk assessment information for the child is received from the terminal device of a personnel of an organization having received the abuse notice, the inputted data is stored in a storage device; a step (B) in which the data of the input item out of the plurality of items is inputted to a first trained model having been trained regarding past abuse cases and a first index value representing the probability of the child being taken into temporary custody is identified; and a step (C) in which the first index value is outputted to the terminal device of the personnel of the organization.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to information processing technology used in child consultation centers and the like. [Background technology]

[0002] In recent years, child abuse has come to be recognized as a major social problem, and the workload of welfare facilities such as child consultation centers has increased dramatically. As a result, staff at child consultation centers and other facilities are overwhelmed with work, find it difficult to improve their expertise, are unable to effectively cooperate with other institutions, and are forced to make decisions that rely solely on experience and intuition.

[0003] Until now, systems have been developed that mainly involve paper-based work, and some systems even have the ability to determine the level of urgency by answering predetermined questions; however, these systems do not provide sufficient support for the decision-making of staff at child consultation centers and other facilities. [Prior art documents] [Non-patent literature]

[0004] [Non-Patent Document 1] "FUJITSU Local Government Solutions Comprehensive Welfare Child Consultation System", [Online] [Retrieved March 14, 2019] Internet <URL:http: / / www.fujitsu.com / jp / group / fip / solutions / industry-solutions / public-sector-solution / childconsul / > Summary of the Invention [Problem to be solved by the invention]

[0005] Therefore, one object of the present invention is to provide a novel technique for supporting the decision-making of staff at child consultation centers and the like. [Means for solving the problem]

[0006] An information processing method according to a first aspect of the present invention includes the steps of: (A) receiving, from a terminal device of an employee of an organization that accepted the abuse report, data entered for at least one of a first plurality of items, including items regarding information about the child related to the abuse report and items regarding risk assessment information of the child, storing the entered data in a storage device; (B) inputting data for items that have already been entered among the first plurality of items into a first trained model of past abuse cases to identify a first index value that represents the likelihood of temporary protection occurring for the child; and (C) outputting the first index value to the terminal device of the employee of the organization. An information processing method according to a second aspect of the present invention includes the steps of: (A) receiving, from a terminal device of an employee of an organization that accepted the abuse report, data entered for at least one of a first plurality of items, including items regarding information about the child related to the abuse report and items regarding risk assessment information of the child, storing the entered data in a storage device; (B) inputting data for items that have already been entered among the first plurality of items into at least one of a first trained model of past abuse cases and a probabilistic model based on past abuse cases, to identify at least one of a first index value indicating the likelihood of temporary protection occurring for the child and a second index value that differs from the first index value and indicates the degree of risk if the report is overlooked; and (C) outputting at least one of the first index value and the second index value to the terminal device of the employee of the organization. An information processing method according to a third aspect of the present invention includes the steps of: (A) receiving data entered for at least one of a first plurality of items, including items regarding information about the child related to the abuse report and items regarding risk assessment information of the child, from a terminal device of an employee of an organization that accepted the abuse report, storing the entered data in a storage device; (B) inputting data for items that have already been entered among the first plurality of items into a probabilistic model based on past abuse cases, and identifying an index value that represents the degree of risk if the report is overlooked for each type of response to the child; and (C) outputting the second index value to the terminal device of the employee of the organization. Effect of the Invention

[0007] On the one hand, it will enable appropriate support for decision-making by staff at child consultation centers and other facilities. [Brief description of the drawings]

[0008] [Figure 1] FIG. 1 is a diagram showing an overview of a system according to an embodiment of the present invention. [Diagram 2] FIG. 2 is a diagram illustrating an example of data stored in a database. [Diagram 3] FIG. 3 is a diagram illustrating an example of a business flow. [Figure 4] FIG. 4 is a diagram showing a process flow in phase A of the business. [Diagram 5] FIG. 5 is a diagram showing an example of an input screen on the terminal device. [Figure 6] FIG. 6 is a diagram showing an example of a simulation result display screen. [Figure 7] FIG. 7 is a diagram showing an example of the decision-making screen. [Figure 8] FIG. 8 is a diagram showing a flow of processing related to total risk calculation. [Figure 9] FIG. 9 is a diagram showing a flow of processing related to prediction of the number of response days. [Figure 10] FIG. 10 is a diagram illustrating a flow of the simulation result generation process. [Figure 11] FIG. 11 is a diagram showing a process flow in phase B of the business. [Figure 12] FIG. 12 is a diagram showing a process flow in phase C of the business. [Figure 13] FIG. 13 is a diagram showing an example of an input screen in phase C of a business operation. [Figure 14] FIG. 14 is a block diagram of a computer device. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0009] An overview of a system according to an embodiment of the present invention is shown in Fig. 1. In this embodiment, each staff member at a child consultation center or the like performs work using a terminal device 2000 such as a tablet. This allows work to be performed not only at a business such as a child consultation center, but also in various locations such as the home of a child (also called a child) who has been notified of abuse.

[0010] The terminal device 2000 performs processing while communicating with the information processing device 1000 via various communication media. For example, a dedicated application program may be installed in the terminal device 2000, and the terminal device 2000 may execute the dedicated application program to execute processing in cooperation with the information processing device 1000, or a general-purpose web browser may execute processing in cooperation with the information processing device 1000. The information processing device 1000 may be a single computer or an information processing system including multiple computers. The information processing device 1000 may also be one or multiple virtual servers realized by a cloud or the like.

[0011] In order to realize the functions according to this embodiment, the information processing device 1000 has a processing unit 1010, a database 1020, a prediction unit 1030, and a learning unit 1040.

[0012] The processing unit 1010 serves as an interface with the terminal device 2000, registers received data in the database 1020, reads data from the database 1020 and transmits it to the terminal device 2000, and has the prediction unit 1030 perform prediction processing of various data based on the data stored in the database 1020, etc.

[0013] The database 1020 accumulates various data on children for whom abuse has been reported. As data relating to this embodiment, for example, data as shown in FIG. 2 is stored. In the example of FIG. 2, for each child (i.e., for each child ID), data on multiple items related to information on the child related to the abuse report and risk assessment information of the child, a person in charge ID, total risk, predicted response days for each type of response to the child, severity for each type of response to the child, recurrence rate for each type of response to the child, latent class, whether temporary protection was implemented, actual response days (number of days elapsed from the date and time of receipt of the report), whether recurrence occurred, content of decision-making, whether it corresponds to a serious case cluster, etc. are stored. In addition, for example, the code of the child consultation center and staff data related to the years of experience of staff at the child consultation center, etc. may also be stored for each child. In addition, school age category, time period of notification including the date and time of notification, period until recurrence (period from the date and time of previous termination to the date and time of recurrence), etc. may be stored.

[0014] The prediction unit 1030 has a total risk prediction unit 1031, a number of days to respond prediction unit 1032, a severity prediction unit 1033, a recurrence prediction unit 1034, and a class prediction unit 1035. The learning unit 1040 has a first learning unit 1041 that performs learning processing on the total risk prediction unit 1031 and the number of days to respond prediction unit 1032, a second learning unit 1042 that performs learning processing on the severity prediction unit 1033 and the recurrence prediction unit, and a third learning unit 1043 that performs learning processing on the class prediction unit 1035.

[0015] The comprehensive risk prediction unit 1031 is a module that includes a trained model based on machine learning and predicts an index value that indicates the likelihood of a child being placed in temporary protection from data on the following items: In other words, it outputs the probability (a number between 0% and 100%) of a child being placed in temporary protection for a certain report based on past abuse case patterns.

[0016] More specifically, the comprehensive risk prediction unit 1031 includes a trained model based on gradient boosting. Boosting is one of the modeling techniques of machine learning that uses multiple weak learners to construct a new, stronger learner with as high versatility as possible. In many cases, multiple conditional branching algorithms called decision trees are used in combination. More precisely, when creating a new decision tree as boosting, a decision tree algorithm is adopted that uses the results of the previous decision tree and minimizes the error between the actual measured value and the predicted value. That is, gradient boosting is a method that uses gradient descent when minimizing the error between the actual correct answer and the prediction as a loss function. More specifically, in this embodiment, Extreme Gradient Boosting (Tianqi Chen and Carlos Guestrin. Xgboost: A scalable tree boosting system. In Proceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining, pages 785-794. ACM, 2016) is used. The specific processing contents are well known, so a detailed description will be omitted.

[0017] The first learning unit 1041 learns to predict whether or not temporary protection will be implemented based on data of specific items in past abuse cases, and generates a trained model that outputs a predicted score for whether or not temporary protection will be implemented. Furthermore, the first learning unit 1041 generates a function that represents the relationship between the predicted score and the probability that temporary protection will occur, for example, by logistic curve approximation. The first learning unit 1041 also performs logistic curve approximation. In other words, the comprehensive risk prediction unit 1031 includes a trained model that outputs a predicted score, and a function that converts the predicted score into the probability that temporary protection will occur.

[0018] The response days prediction unit 1032 is a trained model based on machine learning, and is a module that outputs the number of days predicted to be required to respond to a child for each type of response to the child, based on data on the following items. That is, it predicts the number of days from the date of receipt of a notice to the conclusion based on past abuse case patterns. More specifically, the response days prediction unit 1032 is a trained model based on gradient boosting. Note that in addition to simply predicting the number of response days, there are also cases where the predicted number of response days is categorized into a predetermined range of days and outputted.

[0019] In addition, the first learning unit 1041 also performs learning processing on the response days prediction unit 1031, together with the overall risk prediction unit 1031 which includes a trained model based on the same gradient boosting.

[0020] The severity prediction unit 1033 is a module that predicts severity, which is an index value indicating the probability that a case belongs to a serious case cluster, for each type of response to a child, from data on the following items. That is, it outputs the probability that a certain report (or a child related to the report) belongs to a serious case cluster, based on past abuse case patterns. The severity can also be said to be an index value indicating the risk of overlooking a report, and is also called the risk of overlooking a report.

[0021] More specifically, the severity prediction unit 1033 includes a probabilistic model based on a Bayesian network. A Bayesian network is one of the graphical models that hierarchically depicts probabilistic causal relationships using nodes that represent phenomena (e.g., types of abuse, age categories, etc.) and their connection relationships. For observed data (=nodes), the most applicable and explainable model is exploratory based on the conditional probability of the observed value from the probabilistic causal relationships obtained from Bayes' theorem. During modeling, the relationship between parent nodes and child nodes in a hierarchical structure is depicted. In a Bayesian network, a model that maximizes the conditional probability that a child node is observed under the conditions in which a parent node is observed is adopted. This makes it possible to perform a simulation (probabilistic causal inference) in which, for example, when a child's data is input, it is compared with past data and calculated, and the probability of belonging to a severe cluster is 80% based on the currently known information, even in a situation with high uncertainty.

[0022] For more specific details of the processing, see, for example, J. Pearl, “Bayesian networks: A model of self-activated memory for evidential reasoning.” Proc. 7th Conf. Cogn. Sci. Soc., pp. 329-334, 1985. The specific processing content is well known, so a detailed explanation will be omitted.

[0023] Using a Bayesian network in this way to perform probabilistic causal inference is extremely effective in making predictions that are probabilistically explainable in administrative frameworks such as child consultation centers, where accountability is strongly expected.

[0024] The second learning unit 1033 classifies past abuse cases into multiple clusters, such as a serious case cluster and a minor injury cluster, based on data of predetermined items in the past abuse cases using probabilistic latent semantic analysis (pLSA). In this case, instead of just classifying into one of the clusters, the probability of belonging to each cluster is also obtained at the same time. Then, past abuse cases for which the cluster with the highest probability is the serious case cluster are deemed to belong to the serious case cluster.

[0025] Note that pLSA is a method of clustering by dimensionality compression, and is a type of soft clustering in which the probability of belonging to each cluster is calculated. For details, see, for example, Hofmann, T. (1999) "Probabilistic latent semantic analysis." Proceedings of the Fifteenth conference on Uncertainty in artificial intelligence. Morgan Kaufmann Publishers Inc. The specific processing content is well known, so a detailed explanation will be omitted.

[0026] Then, the second learning unit 1033 executes a process of constructing a probabilistic model using a Bayesian network that predicts the probability of belonging to the severe class from data on specified items in past abuse cases.

[0027] The recurrence prediction unit 1034 is a module that predicts an index value that indicates the likelihood of a child being reported again for each type of child response from data on the following items: In other words, it outputs the probability (a number between 0% and 100%) that a child related to a certain report will be reported again based on past abuse case patterns.

[0028] More specifically, the recurrence probability predicting unit 1034 includes a probability model based on a Bayesian network. In this embodiment, the recurrence probability predicting unit 1034 also adopts the severity as an input item.

[0029] The second learning unit 1042 executes a process of constructing a probabilistic model using a Bayesian network to predict the recurrence rate from data on specific items in past abuse cases. This makes it possible to perform a simulation (probabilistic causal inference) in which, for example, by inputting a child's data, the data is compared with past data and calculated, and a future recurrence rate of 80% is predicted based on the currently known information, even in a highly uncertain situation.

[0030] The class prediction unit 1035 is a module that predicts to which class (latent class) the child related to the report belongs, based on data of the items described below.

[0031] The third learning unit 1043 clusters past abuse cases based on the predetermined items in the past abuse cases using latent class analysis. At this time, for example, the class with the highest belonging probability is identified as the belonging class (latent class). Then, the third learning unit 1043 learns based on gradient boosting (e.g., Xgboost) so as to predict the belonging class from the predetermined items in the past abuse cases, and generates a learned model.

[0032] Like pLSA, latent class analysis is a type of soft clustering. For details of latent class analysis, see, for example, Dayton, C. Mitchell. 1998. Latent Class Scaling Analysis. Thousand Oaks, CA: SAGE Publications. Hagenaars, Jacques A. and Allan L. McCutcheon, eds. 2002. Applied Latent Class Analysis. Cambridge: Cambridge University Press. For the EM algorithm used in latent class analysis, see McLachlan, Geoffrey J. and Thriyambakam Krishnan. 1997. The EM Algorithm and Extensions. New York: John Wiley & Sons. The specific processing content is well known, so a detailed explanation will be omitted.

[0033] The learning unit 1040 performs the learning process, for example, periodically or at an instructed timing.

[0034] Next, the workflow at a child consultation center or the like will be explained using Figure 3. First, a report of child abuse is received at the child consultation center or the like (step S1). In response to this report, an emergency reception conference is held (step S3). Steps S1 and S3 are included in Phase A, in which basic information is entered by staff at the child consultation center or the like, and various information to support overall risks and decision-making is presented to the staff at the child consultation center or the like.

[0035] After that, an emergency dispatch is made to confirm the presence of the child (Step S5). Step S5 is included in Phase B, in which the staff of child consultation centers and other facilities are presented with comprehensive risk information and various information to support their decision-making, and then the staff enters their decision regarding the child that was notified.

[0036] After that, the facts are confirmed with the guardian and related contents are investigated (Step S7). Step S7 is included in Phase C, in which the contents of the investigation are entered. Note that even while the contents of the investigation are being entered, various information is presented to support decision-making.

[0037] In child consultation centers and the like, work generally proceeds according to this workflow, but there are cases where it does not always follow this workflow. In any case, when staff at child consultation centers and the like enter the data of children who have been notified, the system supports them in making appropriate decisions about the children by making it easier for them to understand the overall risks and other risks of the children.

[0038] Next, the processing performed in phase A will be described with reference to FIGS.

[0039] The terminal device 2000 accepts an input of basic information about a child from a staff member of a child consultation center or the like, and transmits the information to the information processing device 1000 (FIG. 4: step S11). Upon receiving the basic information about the child from the terminal device 2000, the information processing device 1000 stores the information in the database 1020.

[0040] The basic information of the child is information about the child himself / herself that is recorded when an abuse report is received, and includes the child's name, sex, relationship, nationality, date of birth, age, information about siblings, whether the mother is a designated pregnant woman (a pregnant woman in an environment where she requires public support for child rearing under the Child Welfare Act from the time of pregnancy), place of residence, etc. Of this, for example, the sex, age, and whether the mother is a designated pregnant woman are used as inputs to the prediction unit 1030.

[0041] Furthermore, the terminal device 2000 accepts input of information on notification acceptance / urgent acceptance conference from staff of a child consultation center or the like, and transmits it to the information processing method 1000 (step S13). Upon receiving the information on notification acceptance / urgent acceptance conference from the terminal device 2000, the information processing device 1000 stores it in the database 1020.

[0042] FIG. 5 shows an example of a screen displayed on the terminal device 2000 for inputting information about the notification acceptance / urgent acceptance conference.

[0043] The information on report acceptance and emergency acceptance conference is information on the case that is input when an abuse report is accepted or when information is collected from related institutions, and includes the status (new / re-accepted / re-notification), the date and time of the report, the circumstances of the report acceptance (type of acceptance route), the main abuser, the type of abuse, the date and time of the emergency acceptance conference, etc. Of this, for example, the status, the date and time of the report, the circumstances of the report acceptance, the main abuser, and the type of abuse are used as inputs to the prediction unit 1030.

[0044] As shown in the upper part of Figure 5, the following information will be entered in sequence: 2. Emergency deployment, 3. Temporary protection consideration items, 4. Limit rule items, 5. Decision-making and safety responsibility, and 6. Implementation items.

[0045] In this embodiment, even during such an input stage, the total risk display 50 is made. That is, the total risk is calculated and displayed sequentially (step S31). This step S31 is executed at any stage of steps S11 to S21. However, it may not be executed at some stages.

[0046] For this reason, the processing unit 1010 makes the overall risk prediction unit 1031 predict the overall risk using data for already-entered items among the predetermined items used for prediction, and default data for items not yet entered, and transmits the predicted overall risk data to the terminal device 2000, thereby causing the terminal device 2000 to sequentially display it. This allows staff at child consultation centers and the like to recognize the likelihood of temporary protection occurring each time data is entered.

[0047] For example, the data entered in steps S11 and S13 is information about the child for whom the abuse has been reported.

[0048] Furthermore, the terminal device 2000 accepts input of emergency dispatch information from staff of a child consultation center or the like, and transmits it to the information processing device 1000 (step S15). When the information processing device 1000 receives the emergency dispatch information from the terminal device 2000, it stores it in the database 1020. The emergency dispatch information is a check item when considering an emergency dispatch for the purpose of confirming the safety of a child, and includes items such as whether or not there has been a history of abuse in the past (or unknown), whether or not the child feels anxious or scared about returning home (or unknown), and whether or not the child currently has a bruise (or unknown). Note that the bruise may include information such as an image taken by a camera or the like of the terminal device 2000, and the type of wound. For example, these data are used as input to the prediction unit 1030.

[0049] Furthermore, the terminal device 2000 accepts input of temporary protection review items from staff of a child consultation center or the like, and transmits the input to the information processing device 1000 (step S17). Upon receiving information on the temporary protection review items from the terminal device 2000, the information processing device 1000 stores the information in the database 1020.

[0050] The information on the items for temporary protection consideration is a check item for considering temporary protection, which is a compulsory intervention to temporarily separate a child from an abusive environment, and includes items such as the presence or absence (or unknown) of bruises on the head, face, or abdomen, the presence or absence (or unknown) of fractures, burns, SBS (head trauma), MSBP (including suspicion), the presence or absence (or unknown) of sexual abuse (including suspicion), the presence or absence (or unknown) of medical (healthcare) neglect or medical abuse such as MSBP, etc. For example, these data are used as inputs to the prediction unit 1030.

[0051] Furthermore, the terminal device 2000 accepts input of limit rule information from staff of a child consultation center or the like, and transmits it to the information processing device 1000 (step S19). Upon receiving the limit rule information from the terminal device 2000, the information processing device 1000 stores it in the database 1020.

[0052] The limit rule information is information for considering the setting of a limit rule that promises a parent and a child to be forced to intervene if the parent breaks the promise. For example, the limit rule information includes items such as whether or not there is a situation in which the child's safety cannot be confirmed, such as canceling a designated interview without notice or refusing a home visit, despite the parent being informed in advance of the need for guidance and support during legal measures. For example, these data are used as inputs to the prediction unit 1030.

[0053] The data received in steps S15 to S19 is child risk assessment information.

[0054] Even during this input stage, if instructed by staff at the child consultation center or the like, a simulation process is executed as needed, and the simulation results are displayed (step S35).

[0055] An example of a display screen of the simulation results is shown in Fig. 6. In the example of Fig. 6, the types of care for children that can be selected are at-home support (selection button 61), at-home with limit rules (selection button 62), and temporary protection (selection button 63).

[0056] That is, in this embodiment, the types of responses to children are at home, at home with limit rules, and temporary protection. Simply at home indicates a situation where the risk is extremely low or corresponds to a case that does not qualify as abuse (for example, a simple cry is reported and not abuse), and there is no particular guidance, support, or legal measures. At home with limit rules indicates a situation where it is determined that there is a risk of abuse occurring again in the future, but that the risk is not so high that protection is required. For this reason, a framework for support is presented, like probation, and guidance measures such as regular interviews and home visits are legally mandatory. If this is violated (limit rules), the risk judgment will be reviewed again, and temporary protection will be reconsidered. Temporary protection indicates a situation where a child is forced to temporarily enter a protective facility, etc.

[0057] Selecting any one of these changes the content of the radar chart display section 65. That is, since the recurrence rate, predicted number of days to respond ("Number of days to respond" in FIG. 6), and severity are calculated for each type of response to the child, the radar chart is also displayed in a different form.

[0058] In the radar chart display section 65, the recurrence rate is shown as points ranging from 0 (origin) to 100% (maximum value of the axis). The severity rate is also shown as points ranging from 0 (origin) to 1 (maximum value of the axis). The predicted number of days to respond is shown as points ranging from 0 days (origin) to 365 days (maximum value of the axis; if it is 366 days or more, it is set to 365 days).

[0059] Staff at child consultation centers, etc., grasp the relative magnitude relationship of these three evaluation index values ​​for each type of response to children, and confirm which type of response will reduce the coverage area on the radar chart. It is preferable for all three evaluation index values ​​to be small, but it is not possible to reduce them all at the same time. For example, even if the severity is low, there may be an event where the recurrence rate or predicted number of days to respond is high. Therefore, the appropriate type of response should be selected while also referring to the overall risk displayed in the overall risk display section 64.

[0060] In this embodiment, past similar cases 67 and comments 66 are displayed below the total risk display section 64 and the radar chart display section 65.

[0061] For past similar cases 67, data on past cases clustered into the same latent class as the latent class of the notification or the child related to the notification is extracted and displayed. For example, data on the latest three past cases is extracted and a part of the data is displayed. For example, by selecting any of the cases, the details of that case are displayed.

[0062] Moreover, the comments 66 include sentences showing characteristics and patterns (for example, patterns with a corresponding probability equal to or greater than a predetermined value) prepared in advance according to the latent class. Such data may be stored in the terminal device 2000 in advance and read out according to the identifier of the latent class, etc. Also, the data may be acquired from the database 1020 of the information processing device 1000.

[0063] Then, the terminal device 2000 receives data for displaying decision making including categories of predicted response days for each type of response to a child from the information processing device 1000, and displays a decision making screen (step S21). At least the categories of predicted response days for each type of response to a child and the overall risk are received from the information processing device 1000.

[0064] For example, a decision-making screen as shown in Fig. 7 is displayed. The example in Fig. 7 includes an at-home selection button 71 including a category display of the predicted number of days to respond when at-home is selected, an at-home selection button 72 with limit rules including a category display of the predicted number of days to respond when at-home with limit rules is selected, a temporary protection selection button 73 including a category display of the predicted number of days to respond when temporary protection is selected, and a total risk display section 74. The total risk display section 74 is the same as the total risk display 50 in Fig. 5.

[0065] In this way, on the decision-making screen, the user can objectively select the type of care to be given to the child by looking at the overall risk and comparing the predicted number of days for care (i.e., the range of the relevant category) for each type of care to be given to the child.

[0066] Next, the process related to the total risk calculation executed in the information processing device 1000 will be described with reference to FIG.

[0067] When the processing unit 1010 of the information processing device 1000 receives new input data from the terminal device 2000, it stores the data in the database 1020 (step S51). In the case of a completely new child, a child ID is issued and a new record is generated in the database 1020, but in the case of a child who is already registered, the new input data is registered in association with the child ID.

[0068] Then, the processing unit 1010 outputs the data of the predetermined items to the overall risk prediction unit 1031 to predict the overall risk and identify the overall risk (step S53). In this embodiment, of the multiple predetermined items related to the information on the child related to the abuse report, the risk assessment information of the child, and the information on how to respond to the child (specifically, the information entered in Phase B), the data of the items that have already been input is used, and for items that have not been input, default data is used to cause the overall risk prediction unit 1031 to predict the overall risk.

[0069] Since some information cannot be researched or is unknown, it may not be possible to input information for all of the multiple specified items. Therefore, for items where data has been input, the input data is used, and for items where no data has been input or the contents are unknown, predetermined default data is used.

[0070] As described above, the overall risk prediction unit 1031 has been trained to predict whether or not temporary protection will be implemented based on specified items of data in past abuse cases, calculates a prediction score for whether or not temporary protection will be implemented, converts the prediction score into the probability that temporary protection will be implemented, and outputs it as the overall risk.

[0071] The processing unit 1010 transmits the total risk predicted by the total risk prediction unit 1031 to the terminal device 2000 (step S55). The processing unit 1010 registers the total risk in the database 1020.

[0072] This will allow staff at child consultation centers and other facilities to grasp the likelihood of temporary protection occurring on an ongoing basis, even while they are in the process of entering information.

[0073] Next, a process related to prediction of the number of days to be covered, which is executed in the information processing device 1000 before step S21 in FIG. 4, will be described with reference to FIG.

[0074] The processing unit 1010 reads data of specified items from the database 1020 and outputs it to the response days prediction unit 1032, causes the response days prediction unit 1032 to execute processing to predict the number of response days for each response type, and identifies the predicted result of the number of response days for each response type (step S61).

[0075] In this embodiment, of the multiple specified items related to information about the child related to the abuse report, risk assessment information about the child, and information about how to respond to the child (specifically, information entered in Phase B), data for items that have already been entered is used, and default data is used for items that have not been entered, to cause the response days prediction unit 1032 to predict the number of response days.

[0076] The response days prediction unit 1032 assumes a decision to "stay at home" as information regarding how to treat a child, and predicts the number of response days in the case of "stay at home" from data of specified items. Also, assuming a decision to "stay at home with limit rules", the unit predicts the number of response days in the case of "stay at home with limit rules" from data of specified items. Furthermore, assuming a decision to "temporarily protect", the unit predicts the number of response days in the case of "temporarily protect" from data of specified items.

[0077] Also, the number of days to be handled prediction unit 1032 predicts a specific number of days (integer) of 0 or more. In the terminal device 2000, the specific number of days may be presented as it is, but there is a problem of how to grasp the specific number of days. That is, there is a problem that it is difficult to grasp the level of effort required from the number of days alone. Therefore, in this embodiment, for example, it is divided into six categories: category A of less than 14 days, category B of 14 days or more and less than 30 days, category C of 30 days or more and less than 90 days, category D of 90 days or more and less than 180 days, category E of 180 days or more and less than 365 days, and category F of 365 days or more. The number of categories is arbitrary.

[0078] Therefore, the response days prediction unit 1032 determines a category corresponding to the predicted response days for each response type (step S63). The response days prediction unit 1032 outputs at least one of the predicted response days and the category to the processing unit 1010. The processing unit 1010 also stores at least one of the predicted response days and the category in the database 1020. Note that the processing unit 1010 may execute this step.

[0079] Then, the processing unit 1010 generates data for displaying decision making including a category of the predicted number of days for each action type, and transmits it to the terminal device 2000 (step S65). As described above, the terminal device 2000 displays a screen such as that shown in Fig. 7. The contents of the total risk display section 74 included in the example screen of Fig. 7 will be displayed separately by processing such as that shown in Fig. 8.

[0080] In this way, staff at child consultation centers and other facilities can easily identify the type of treatment that is expected to shorten the number of days required for treatment. In other words, support for decision-making by staff at child consultation centers and other facilities can be provided appropriately.

[0081] Next, the process performed when a staff member of a child consultation center or the like instructs to display the simulation results will be described with reference to FIG.

[0082] The processing unit 1010 of the information processing device 1000 outputs data of predetermined items in the database 1020 to the overall risk prediction unit 1031 to predict the overall risk and identify the overall risk (step S71). The processing of this step is the same as step S53 in Fig. 8, and causes the overall risk prediction unit 1031 to predict the overall risk using data for items that have already been input among the multiple predetermined items, and default data for items that have not been input. The processing unit 1010 stores the overall risk in the database 1020.

[0083] The processing unit 1010 also outputs data of predetermined items in the database 1010 to the response days prediction unit 1032 to predict the number of response days for each response type and specify the predicted number of response days (step S73). The processing of this step is the same as step S61 in Fig. 9, and among the multiple predetermined items, for data of items that have already been input, the data is used, and for data of items that have not been input, default data is used to cause the response days prediction unit 1032 to predict the number of response days. The processing unit 1010 stores the predicted number of response days in the database 1020.

[0084] Furthermore, the processing unit 1010 outputs data of predetermined items in the database 1010 to the severity prediction unit 1033 to have the unit predict the severity for each response type and identify the severity (step S75).

[0085] The severity prediction unit 1033 predicts the severity using data from a number of predetermined items related to information about the child related to the abuse report, risk assessment information about the child, information about how to deal with the child (specifically, information entered in Phase B), and information about the person in charge of the child, for items that have already been entered, and default data for items that have not been entered. The processing unit 1010 stores the severity in the database 1020.

[0086] The information about the person in charge of the child includes, for example, the code of the child consultation center to which the person in charge belongs, the number of years of experience of the person in charge, etc. The severity prediction unit 1033 may use items different from those used in the overall risk prediction unit 1031 and the response period prediction unit 1032. For example, additional items such as the child's school age category, the time period of notification, and the period until recurrence may be used.

[0087] Moreover, the processing unit 1010 outputs data of predetermined items in the database 1010 to the recurrence probability predicting unit 1034 to predict the recurrence probability for each treatment type and identify the recurrence probability (step S77).

[0088] The recurrence prediction unit 1033 predicts the degree of recurrence using the data for items that have already been input, default data for items that have not been input, and the severity, out of a number of predetermined items related to information about the child related to the abuse report, risk assessment information about the child, information about how to deal with the child (specifically, information entered in Phase B), and information about the person in charge of the child. The processing unit 1010 stores the recurrence degree in the database 1020.

[0089] The recurrence probability prediction unit 1034 may also use items different from those of the overall risk prediction unit 1031 and the response period prediction unit 1032. For example, additional items such as the child's school age category, the time of notification, and the period until recurrence may be used. As described above, the severity is also used.

[0090] Although the severity prediction unit 1033 and the recurrence prediction unit 1034 use items related to information about the person in charge as input, other prediction units may also use such items if a sufficient number of past cases are available. In other words, since the models used in the severity prediction unit 1033 and the recurrence prediction unit 1034 are capable of guaranteeing uncertainty, information about the person in charge can be effectively used even when the number of past cases is small.

[0091] Then, the processing unit 1010 outputs data of predetermined items in the database 1010 to the class prediction unit 1035 to predict a latent class, identifies the latent class, and extracts data of similar cases from the latent class from the database 1020 (step S79). The class prediction unit 1035 predicts a latent class using data of already-input items and default data of not-input items among a plurality of predetermined items related to information on the child related to the abuse report and the risk assessment information of the child. The processing unit 1010 stores the identifier of the latent class in the database 1020. The processing unit 1010 also identifies past cases in which the same latent class identifier is registered as similar cases, and extracts data of the similar cases from the database 1020.

[0092] In addition, even if information about a child related to a report of abuse or an item included in a child's risk assessment information is included, there may be some items that are not adopted as input by the prediction unit.

[0093] Then, the processing unit 1010 generates data for displaying the simulation result including a radar chart of the total risk, the predicted number of days to respond, the severity, and the recurrence rate, and data on similar cases, and transmits the data to the terminal device 2000 (step S81).

[0094] The processing unit 1010 may generate and transmit data for a screen configuration as shown in Fig. 6, or may transmit the total risk, predicted number of days to respond, severity, recurrence rate, data on similar cases, latent class identifiers, and the like to the terminal device 2000 and render a screen configuration as shown in Fig. 6 on the terminal device 2000. For the predicted number of days to respond, severity, and recurrence rate, values ​​equivalent to the scale of a radar chart may be calculated and transmitted to the terminal device 2000.

[0095] In addition, in steps S73 to S75, the predicted number of days to be taken, the severity, and the recurrence rate are predicted for each type of response, but predictions may be made for the type of response related to the instruction each time in response to an instruction from the terminal device 2000. For example, the predicted number of days to be taken, the severity, and the recurrence rate for the default at-home may be predicted at first, and when temporary protection is instructed, the predicted number of days to be taken, the severity, and the recurrence rate for the temporary protection may be predicted.

[0096] By carrying out the above-mentioned processing, it becomes possible to visually display in an easy-to-understand manner the overall risk associated with this notification and how the type of decision-making will affect the predicted number of days to respond, the severity, and the likelihood of recurrence, thereby enabling staff at child consultation centers and other facilities to provide effective support.

[0097] Next, the process in phase B will be described with reference to FIG.

[0098] The terminal device 2000 accepts a selection of one of the types of response by a staff member of a child consultation center or the like as a decision-making input in the decision-making display shown in Fig. 7, for example, and transmits the selection to the information processing device 1000 (step S101). The processing unit 1010 of the information processing device 1000 stores the received type of response in the database 1020.

[0099] At this time, if temporary protection is selected as the response type, the process proceeds to step S107 (step S103: Yes route).On the other hand, if at-home with limit and at-home are selected (step S103: No route), the terminal device 2000 prompts the staff of the child consultation center or the like to input a reason, accepts the input of the reason, and transmits the content of the reason to the information processing device 1000 (step S105).

[0100] In this embodiment, the reason can be selected from either "suspected child" or "not abused." In the former case, the user can further input details such as specific circumstances as to whether the child is safe and specific circumstances as to why the urgency is low. In the latter case, the user can further input details such as specific circumstances. These inputs are also sent from the terminal device 2000 to the information processing device 1000 and registered in the database 1020.

[0101] Furthermore, the terminal device 2000 accepts input of an action item and transmits data on the action item to the information processing device 1000 (step S107). The processing unit 1010 of the information processing device 1000 registers in the database 1020 the received data on the action item.

[0102] Items to be implemented include the date and time of safety confirmation, whether or not an emergency call was made, whether or not temporary protection was given, the time of protection determination, and the date and time of admission to temporary protection.

[0103] The items input in steps S101 and S103 are included in the information regarding how to deal with the child. For example, the content of the decision-making is used as an input to the prediction unit 1030.

[0104] Once Phase B is completed, the correct data for the learning process will have been obtained, and the data in this notification will be used in the learning process at the next timing of the learning process.

[0105] As shown in Fig. 11, in steps S101 to S107, the process of sequentially calculating and displaying the total risk (step S31) is also performed. Of course, it is also possible to make it so that some of the total risks are not displayed. During this time, if there is an instruction from staff at the child consultation center or the like, a simulation process is executed as needed and the simulation results are displayed (step S35).

[0106] Next, the process in phase C will be described with reference to FIGS.

[0107] The terminal device 2000 accepts input of information on the survey items, and transmits the information to the information processing device 1000 (step S111). The processing unit 1010 of the information processing device 1000 stores the received information on the survey items in the database 1020.

[0108] The survey items include the following data:

[0109] Basic items include whether or not there is a medical or psychiatric diagnosis, whether or not there is an intellectual disability, whether or not there is a physical disability (walking or limbs), whether or not there is a developmental disability (including suspected), whether or not there is school absenteeism, whether or not there is a case of domestic violence (including witnessing violence in front of others), whether or not there is a drug or alcohol use disorder (including suspected), etc. Free description may also be included.

[0110] In addition, the category of confirmation of the child's specific physical and mental condition (including medical evaluation) includes the following items. Free description may also be included. - The specific physical and mental condition of the child reported (in principle, confirmed by direct visual inspection: height, build, complexion, facial expression, communication ability) Regarding visible facts (collection and verification of factual information from the informant, photographing, storing and printing by child welfare officers) - Invisible facts (e.g. cases of head injuries or cases of low body weight of less than 2 SD) (confirmation through medical examinations and tests, forensic opinion (biopsy), questioning: direct interviews with the person and related parties, corroboration of facts through medical and judicial findings), etc.

[0111] Furthermore, in the category of understanding the specific facts and circumstances of the problem, the following items are included. Note that free description may also be included. Confirmation of specific facts of conduct - Confirm facts about who did what, when and where How did the parent-child incident occur? Who was involved and in what relationship did it happen? · Certain circumstances · Whether the reason is known or not Are there any usual episodes about the family or upbringing? Can you guess what daily life is like? -Does the explanation from the person involved touch on the background circumstances? ·Children's affiliations ·Networking information ·Neighborhood relationships Family life and economic situation (general income situation and poverty) · Physical and mental health status (including illness, drinking and smoking habits) ·History of current residence ·History of movement ·Place of birth · Family relationships · Residence · Exchange status ·Important events in life ·Priority of child rearing issues Specific problems with childcare Subjective feelings Objective recognition, etc.

[0112] In addition, the category of parenting ability (not a statement of attitude but a feasible action / prospect of a behavior plan) includes the following items. In addition, free description may be included. · Medical check-up history · Vaccination history · Health care history · Childcare counselors · Supporters · Networking situation and relationships · The caregiver's problem-solving skills · How will they handle the situation next time? -Whether or not (and confirm) whether similar situations have been successfully resolved in the past The caregiver's own self-evaluation, level of confidence and fear Higher fear increases defense, attack and explosiveness -Are you able to objectively grasp the situation? · Current life stress · Problems in the near future · Feelings towards child consultation center staff, feelings felt, etc.

[0113] In addition, the category of "Effectiveness of support from neighbors and relatives: episode level" includes the following items. Note that free description may also be included. - Past interactions with neighbors and relatives as a fait accompli Actual support behavior and degree of help (1) Adjustable? Actual support behavior is a fait accompli, and can we expect help in the future? 2) Is there only future attitude and plans, and concrete actions are unknown? 3) Unclear or the relationship is unstable or distant, so is it difficult to expect anything? -Is there a problem with the relationship itself (1) Possible to adjust? 2) Difficult to adjust? 3) Adjustment required? etc.

[0114] Furthermore, in the category of the effectiveness and expectation of specific assistance from welfare and related organizations, the following items are included. Free description may also be included. -History of using public services to date - Parental health history: Check and evaluate vaccination history, etc. ·Current availability and status of public services Relationship with affiliated institution · Specific consideration · Introduction · Implementation procedure · Contact confirmation · Guidance management by authority setting · Judgment of guidance by child welfare officers, etc.

[0115] Furthermore, the terminal device 2000 accepts input at the time of the end of the survey, and transmits the input data at the time of the end of the survey to the information processing device 1000 (step S113). The processing unit 1010 of the information processing device 1000 stores the received input data in the database 1020.

[0116] For example, input is made on a screen such as that shown in FIG.

[0117] The example in Figure 13 includes a column for selecting the status at the end of the survey (1. Under consideration, 2. In care, 3. Institutional care, 4. Supported at home, 5. In foster care, 6. Closed, 7. Other), an input column for the safety scale at the end of the survey, a column for selecting the category (1. Incoming call, 2. Call, 3. Visit, 4. Visit, 5. Notification), a column for selecting the content of the response and an input column for the number of times, a column for selecting whether to copy the survey information to siblings, a back button, and a registration button.

[0118] The selection field for the response content includes a category selection field for medical diagnostic guidance and other response content, a selection field (subcategory) for the person who will implement the response content (e.g., a doctor), and a selection field (subcategory 2) for the person who is the target of the response (e.g., a child).

[0119] In this embodiment, the safety scale is designed to specify a value using a bar that indicates how safe a child is, where danger is 0 and safety is 100, according to the investigator's assessment. The value specified on the safety scale is used instead of the severity. That is, when calculating the recurrence rate displayed in the simulation results, the severity predicted by the severity prediction unit 1033 is used, but instead, the recurrence rate is calculated using the value specified on the safety scale.

[0120] That is, even at this stage, if there is an instruction from a staff member of the child consultation center or the like, the simulation process is executed and the simulation result is displayed (step S35).

[0121] By using such a system, it will be possible to support staff at child consultation centers and other facilities in various situations. In other words, there will be no forgetting of survey items, and the necessary information will be input, encouraging staff to collect sufficient information. In addition, before a decision is made, the possibility of temporary protection occurring can be displayed as a probability as an overall risk, and the expected number of days for each type of response can be presented, supporting appropriate judgment regarding the need for temporary protection. Furthermore, by visualizing the relative comparison and balance of key indicators such as the expected number of days for response, recurrence rate, and severity for each type of response, it will be possible to further support appropriate judgment.

[0122] Although the embodiment of the present invention has been described above, the present invention is not limited thereto. For example, the functional block configuration or module configuration described above is an example, and a different configuration may be adopted. Specifically, one functional block or module may be divided into multiple blocks, or multiple functional blocks or modules may be integrated into one. Furthermore, the information processing device 1000 and the terminal device 2000 may share functions differently from those described above.

[0123] In addition, as for the processing flow, the order of processing may be changed or multiple steps may be executed in parallel as long as the processing result does not change. For example, a staff member of a child consultation center may change the order of inputs related to steps S13 to S19 in the terminal device 2000. The order of inputs for other steps may also be changed.

[0124] Furthermore, the method employed in the prediction unit 1030 may be replaced by a more preferable method if one is found. Furthermore, the data to be input to the prediction unit 1030 or the data to be input to the terminal device 2000 may be changed to more items, more limited items, or more appropriate items in some cases. Furthermore, the items used by each unit of the prediction unit 1030 may be the same or different. Furthermore, the predicted number of days to respond, the severity, the recurrence rate, and the like may be presented as numerical values ​​themselves.

[0125] The information processing device 1000 described above is a computer device, and as shown in FIG. 14, a memory 2501, a CPU (Central Processing Unit) 2503, a hard disk drive (HDD: Hard Disk Drive) 2505, a display control unit 2507 connected to a display device 2509, a drive device 2513 for a removable disk 2511, an input device 2515, and a communication control unit 2517 for connecting to a network are connected by a bus 2519. The HDD may be a storage device such as a solid state drive (SSD: Solid State Drive). An operating system (OS) and an application program for carrying out the processing in the embodiment of the present invention are stored in the HDD 2505, and are read from the HDD 2505 to the memory 2501 when executed by the CPU 2503. The CPU 2503 controls the display control unit 2507, the communication control unit 2517, and the drive device 2513 according to the processing contents of the application program to perform a predetermined operation. Data during processing is mainly stored in the memory 2501, but may be stored in the HDD 2505. In an embodiment of the present technology, an application program for implementing the above-described processing is stored in a computer-readable removable disk 2511 and distributed, and is installed in the HDD 2505 from the drive device 2513. It may also be installed in the HDD 2505 via a network such as the Internet and a communication control unit 2517. Such a computer device realizes the various functions described above by organic cooperation between hardware such as the above-described CPU 2503 and memory 2501 and programs such as the OS and application programs.

[0126] The data used in executing the above-mentioned processes is stored in a storage device such as the memory 2501 or the HDD 2505, regardless of whether it is data in the middle of processing or the result of processing.

[0127] The above-described embodiment can be summarized as follows.

[0128] The information processing method of this embodiment includes the steps of: (A) receiving input data for at least one of a first plurality of items related to information about a child related to a report of abuse and risk assessment information for the child from a terminal device, storing the input data in a storage device; (B) inputting data for items that have already been input from the first plurality of items into a first trained model of past abuse cases to identify a first index value that represents the likelihood of temporary protection occurring for the child; and (C) outputting the first index value to the terminal device.

[0129] This will make it possible to support the decision-making of staff at child consultation centers, etc. More specifically, whether or not to provide temporary protection for a child related to a report of abuse is a very important decision for child consultation center staff, and if the first index value, which shows the likelihood of this being the case, can be displayed even when only some of the items have been entered, it will enable early and appropriate decision-making.

[0130] In addition, the above information processing method may further include a step of (D) inputting data of already-entered items among a second plurality of items relating to information about the child related to the abuse report and the child's risk assessment information into a first probabilistic model based on past abuse cases, and identifying a second index value representing the probability that the report belongs to a serious case cluster (also referred to as the degree of risk if the report is overlooked) for at least one type of response to the child.

[0131] Even if only some of the items have been entered, it will be useful for staff at child consultation centers and other facilities to be able to obtain a second index value for what would happen if a certain response was taken before deciding what action should be taken with the child in question.

[0132] Furthermore, the above information processing method may further include a step of (E) inputting data of already-entered items among a third plurality of items relating to information about the child related to the abuse report and the child's risk assessment information and the second index value into a second probabilistic model based on past abuse cases, to identify a third index value representing the likelihood that a report against the child will recur for at least any of the types of responses to the child.

[0133] As with the second index value, even if only some of the items have been entered, it is useful for staff at child consultation centers to obtain the third index value when a certain action is taken before deciding what action to take with the child in question. In some cases, the second and third index values ​​are presented together.

[0134] In addition, the above information processing method may further include a step of (F) inputting data of already-input items among a fourth plurality of items related to information about a child related to a report of abuse and risk assessment information for the child into a second trained model of past abuse cases to identify the number of days predicted to be required to respond to the child for at least any of the types of responses to the child, and a step of (G) transmitting data to the terminal device for displaying a relative comparison of the number of days and the second index value and the third index value for at least any of the types of responses to the child.

[0135] As with the second and third index values, even if only some of the items have been entered, it will be useful for staff at child consultation centers and other facilities to obtain the number of days it would take to respond if a certain response is taken before deciding what action should be taken with respect to the child in question.

[0136] Furthermore, it is preferable that the number of days to respond and the second and third index values ​​are all small values, but it is impossible to satisfy them simultaneously. Therefore, it is also useful for decision making to see the balance between them. The display for relative comparison is, for example, a radar chart, but is not limited to this, and may be a graph such as a bar graph in which the maximum value is scaled to the same length. Furthermore, the first index value may be presented at the same time. Also, a display for relative comparison of at least two of the number of days to respond and the second and third index values ​​may be performed. Furthermore, any combination of the number of days to respond and the first to third index values ​​may be presented.

[0137] Furthermore, the information processing method may be configured such that, when a value associated with the second index value is received from the terminal device (H), the third index value is specified by using the value instead of the second index value. For example, a value (e.g., severity) equivalent to the second index value may be determined by an investigation by staff of a child consultation center or the like, and in such a case, the value is utilized.

[0138] Furthermore, the information processing method may further include the step of (I) identifying the class to which the report applies, reading out data on other abuse cases that belong to that class, and transmitting the data to the terminal device, thereby making it possible to refer to the contents of similar past cases.

[0139] In addition, in the step of identifying the first index value described above, the first index value may be identified each time data relating to input of any of the first plurality of items is received. Then, in the step of outputting the first index value to the terminal device, the first index value may be output to the terminal device so as to be displayed on an input screen of at least a part of the first plurality of items in the terminal device. The first index value is updated each time data is input into a predetermined item, making it easier for staff at child consultation centers and the like to objectively grasp the degree of urgency of the notification.

[0140] A program for causing a computer to execute the information processing method described above can be created, and the program is stored in various storage media.

[0141] Furthermore, the information processing device that executes the information processing method described above may be realized by a single computer or may be realized by multiple computers, and these will collectively be referred to as an information processing system or simply a system. [Explanation of symbols]

[0142] 1000 Information processing device 1010 Processing section 1020 Database 1030 Prediction Department 1040 Learning Department

Claims

1. receiving input data for at least one of a plurality of items including items regarding information about the child related to the abuse report and items regarding risk assessment information for the child from a terminal device of an employee of the organization that accepted the abuse report, storing the input data in a storage device; inputting data of a first of the plurality of items into a first trained model of past abuse cases to identify a first index value representing the likelihood of temporary protection occurring for the child; outputting the first index value to a terminal device of an employee of the organization; inputting data of already-input items among a second plurality of items among the plurality of items into a second trained model of past abuse cases, and identifying the number of days predicted to be required for each type of response to the child; outputting data on the number of days to a terminal device of an employee of the organization; A computer-implemented information processing method, comprising:

2. receiving input data for at least one of a plurality of items including items regarding information about the child related to the abuse report and items regarding risk assessment information for the child from a terminal device of an employee of the organization that accepted the abuse report, storing the input data in a storage device; inputting data of a first of the plurality of items into a first trained model of past abuse cases to identify a first index value representing the likelihood of temporary protection occurring for the child; outputting the first index value to a terminal device of an employee of the organization; inputting data of already-input items among a second plurality of items among the plurality of items into a second trained model of past abuse cases, and identifying the number of days predicted to be required for each type of response to the child; outputting data on the number of days to a terminal device of an employee of the organization; A program for causing a computer to execute the above.

3. a means for receiving input data for at least one of a plurality of items including items regarding information about the child related to the abuse report and items regarding risk assessment information for the child from a terminal device of an employee of an organization that has received the abuse report, and storing the input data in a storage device; a means for inputting data of a first plurality of items among the plurality of items into a first trained model of past abuse cases to identify a first index value representing the likelihood of temporary protection occurring for the child; a means for inputting data of already-input items among a second plurality of items among the plurality of items into a second trained model of past abuse cases, and for each type of response to the child, specifying the number of days predicted to be required to respond to the child; means for outputting data relating to the first index value and the number of days to a terminal device of an employee of the organization; An information processing system having the above configuration.

4. receiving input data for at least one of a plurality of items including items regarding information about the child related to the abuse report and items regarding risk assessment information for the child from a terminal device of an employee of the organization that accepted the abuse report, storing the input data in a storage device; inputting data of the previously input items among the plurality of items into a first trained model of past abuse cases to identify a first index value representing the likelihood of temporary protection occurring for the child; outputting the first index value to a terminal device of an employee of the organization; An information processing method comprising: In the step of identifying the first index value, identifying the first index value each time data relating to any one of the plurality of items is received; In the outputting step, outputting the first index value to the terminal device so as to display the first index value on an input screen for at least a part of the plurality of items on the terminal device; A computer implemented information processing method.

5. receiving input data for at least one of a plurality of items including items regarding information about the child related to the abuse report and items regarding risk assessment information for the child from a terminal device of an employee of the organization that accepted the abuse report, storing the input data in a storage device; inputting data of the previously input items among the plurality of items into a first trained model of past abuse cases to identify a first index value representing the likelihood of temporary protection occurring for the child; outputting the first index value to a terminal device of an employee of the organization; The computer executes the following: In the step of identifying the first index value, identifying the first index value each time data relating to any one of the plurality of items is received; In the outputting step, The first index value is output to the terminal device so as to be displayed on an input screen for at least a part of the plurality of items in the terminal device. program.

6. a means for receiving input data for at least one of a plurality of items including items regarding information about the child related to the abuse report and items regarding risk assessment information for the child from a terminal device of an employee of an organization that has received the abuse report, and storing the input data in a storage device; a means for inputting data of already-input items among the plurality of items into a first trained model of past abuse cases to identify a first index value representing the likelihood of temporary protection occurring for the child; means for outputting the first index value to a terminal device of an employee of the organization; having The means for identifying the first index value comprises: identifying the first index value each time data relating to any one of the plurality of items is received; The output means includes: The first index value is output to the terminal device so as to be displayed on an input screen for at least a part of the plurality of items in the terminal device. Information processing system.

Images