Danger prediction training support device and danger prediction training support method

Abstract

To support danger prediction training in which each person considers danger in work.SOLUTION: The danger prediction training support device 1 includes a model sentence database 12 that stores a model sentence of a danger prediction activity including a work procedure item, a risk factor item, and a countermeasure item, an input unit 11 that receives an input of a danger prediction activity sentence by a user, and a comparison result display unit 17 that identifiably displays a model sentence similar to the danger prediction activity sentence and a similarity with the danger prediction activity sentence together with the risk factor item and / or the countermeasure item of the danger prediction activity sentence input to the input unit 11.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a danger prediction training support device and a danger prediction training support method. [Background technology]

[0002] In order to safely carry out work at construction sites, maintenance work, and other such work, hazard prediction activities, including risk assessments, are carried out before the work begins. Patent Document 1 describes a system in which users predict hazards from videos of work scenes and rate the predictions. However, because the content and situations of work vary widely, collecting training videos is generally difficult.

[0003] On the other hand, Patent Document 2 describes a system that focuses on risk prediction activities before work and supports the creation of documentation for such activities. In this system, the system presents a group of candidate risk factors and countermeasures according to the work content, and the user selects the appropriate content. However, this system lacks a process for the user to think independently, and is therefore insufficient as risk prediction training. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Publication No. 2017-083664 [Patent Document 2] Patent No. 7369885 Summary of the Invention [Problem to be solved by the invention]

[0005] To ensure the safety of workers and their co-workers, it is important to improve risk prediction abilities and to be able to verbalize risk factors and countermeasures. In response to this, risk prediction activities before work are widely practiced. However, it is particularly difficult for unskilled workers to make up for their own lack of knowledge and perspectives. There is an urgent need to improve the risk prediction abilities of unskilled workers, who are at a high risk of accidents.

[0006] Therefore, an object of the present invention is to support risk prediction training in which each person thinks about the risks involved in their work. [Means for solving the problem]

[0007] In order to solve the above-mentioned problems, the hazard prediction training support device of the present invention is characterized by comprising a model sentence database that stores model sentences for hazard prediction activities including work procedure items, risk factor items, and countermeasure items, an input unit that accepts hazard prediction activity sentences input by a user, and a comparison result display unit that identifiably displays model sentences similar to the hazard prediction activity sentence, as well as the work procedure items and / or risk factor items and / or countermeasure items of the hazard prediction activity sentence input to the input unit, and the degree of similarity with the hazard prediction activity sentence.

[0008] The hazard prediction training support method of the present invention is characterized by comprising a step in which an input unit accepts input of a hazard prediction activity sentence by a user, and a step in which a comparison result display unit displays, in a identifiable manner, a model sentence similar to the hazard prediction activity sentence, and the degree of similarity with the hazard prediction activity sentence, along with the risk factor items and / or countermeasure items of the hazard prediction activity sentence input to the input unit. Other means will be described in the detailed description of the invention. [Effects of the Invention]

[0009] According to the present invention, it is possible to support risk prediction training in which each person thinks about the risks involved in their work. [Brief explanation of the drawings]

[0010] [Figure 1] 1 is a functional configuration diagram of a danger prediction training support device according to an embodiment of the present invention. [Figure 2] FIG. 2 is a hardware configuration diagram of the danger prediction training support device. [Figure 3] 10 is a flowchart of a hazard prediction training support process. [Figure 4] FIG. 10 is a diagram showing an example of a sentence input by a user. [Figure 5] FIG. 10 is a diagram showing an example of a model sentence. [Figure 6] FIG. 10 is a diagram illustrating a similarity calculation process. [Figure 7] FIG. 10 is a diagram illustrating a similarity calculation process. [Figure 8] FIG. 10 is a diagram illustrating a similarity calculation process. [Figure 9] FIG. 10 is a diagram illustrating a similarity calculation process. [Figure 10A] FIG. 10 is a diagram showing a display screen of similarity. [Figure 10B] FIG. 10 is a diagram showing a display screen of similarity. [Figure 10C] FIG. 10 is a diagram showing a display screen of similarity. DETAILED DESCRIPTION OF THE INVENTION

[0011] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

[0012] FIG. 1 is a functional configuration diagram of a danger prediction training support device 1 according to this embodiment. The hazard prediction training support device 1 comprises an input unit 11, a model sentence database 12, a similar line selection unit 13, an item similarity calculation unit 14, a display target extraction unit 15, a sentence similarity calculation unit 16, a comparison result display unit 17, and a threshold setting unit 18. The hazard prediction training support device 1 is a device that compares sentences input for hazard prediction training with model sentences and displays the degree of similarity.

[0013] The input unit 11 accepts input of hazard prediction activity sentences by the user. The input hazard prediction activity sentences include work procedure items, risk factor items, and countermeasure items. The model sentence database 12 stores model sentences for hazard prediction activities, each containing a work procedure item, risk factor item, and countermeasure item on a line-by-line basis. These model sentences are training materials for hazard prediction activities.

[0014] The training materials consist of hazard prediction activity sentences with a work subject, with three items - work procedures, risk factors, and countermeasures - written side by side on each line. The system selects model sentences from the database that are similar to the input hazard prediction activity sentence. When evaluating the similarity with each model sentence, the line with the greatest similarity in work procedures in the input hazard prediction activity sentence is linked as the similar line, and the similarity in risk factors and countermeasures is also calculated. The comparison results screen lists each line of the input hazard prediction activity sentence and its similar lines, and also displays the similarity of the three items mentioned above. Items with low similarity contain information that focuses on a different point of view from the input hazard prediction activity sentence, so the relevant parts are colored to make you aware of any missing information.

[0015] The similar line selection unit 13 selects model sentences similar to the hazard prediction activity sentences input to the input unit 11 from the model sentence database 12. When evaluating the similarity with each model sentence, the similar line selection unit 13 links the line with the greatest similarity in work procedure among the lines of the input hazard prediction activity sentences as the similar line.

[0016] The item similarity calculation unit 14 calculates the similarity between the risk factor items and / or countermeasure items of the danger prediction activity sentence input to the input unit 11 and the model sentence selected by the similar line selection unit. The display target extraction unit 15 extracts the contents of each item of work procedure, risk factor, and countermeasure of the model text that has the largest average value of similarity between work procedure items, risk factor items, and countermeasure items through the calculations of the item similarity calculation unit 14, from each line of the hazard prediction activity text entered into the input unit 11. The text similarity calculation unit 16 calculates the similarity as a text based on the similarity of the items extracted by the display target extraction unit 15.

[0017] The comparison result display unit 17 displays the work procedure items, risk factor items, and countermeasure items of the hazard prediction activity text entered into the input unit 11, along with the extracted contents corresponding to those items by the display target extraction unit 15, and the similarity between the two items, in a manner that allows identification for each item.

[0018] The comparison results screen displays each line of the entered hazard prediction activity text alongside similar lines, and also displays the degree of similarity between work procedure items, risk factor items, and countermeasure items. Items with low similarity contain information that focuses differently from the entered hazard prediction activity text, so the relevant items can be highlighted to make users aware of any missing information.

[0019] The threshold setting unit 18 sets a threshold for the comparison result display unit 17. This allows the threshold setting unit 18 to change the display of the comparison result for items with low similarity, thereby making it possible to give people an awareness.

[0020] FIG. 2 is a hardware configuration diagram of the danger prediction training support device 1. As shown in FIG. The risk prediction training support device 1 includes a CPU (Central Processing Unit) 91, a ROM (Read Only Memory) 92, and a RAM (Random Access Memory) 93. The CPU 91 executes a program (not shown) to realize the functions of the risk prediction training support device 1. The ROM 92 is a non-volatile memory and stores, for example, a BIOS (Basic I / O System), etc. The RAM 93 is a volatile memory and stores data temporarily held by the CPU 91 that executes the program.

[0021] The risk prediction training support device 1 further includes a memory unit 94, a display unit 95, an input unit 96, and a communication interface 97, which are interconnected by a bus 98. The storage unit 94 is a large-capacity storage unit such as a hard disk or SSD (Solid State Drive), and stores programs, data, and the like.

[0022] The display unit 95 is, for example, a liquid crystal display, and displays characters, figures, images, etc. The display unit 95 may be detachable. The input unit 96 is, for example, a transparent touch panel superimposed on the liquid crystal display, and accepts user input operations. The communication interface 97 communicates with, for example, a cloud server. The communication interface 97 may support any communication format, such as a wired NIC (Network Interface Card), a wireless NIC, or a fifth-generation communication module.

[0023] FIG. 3 is a flowchart of the hazard prediction training support process. First, the input unit 11 accepts an input of a danger prediction activity sentence from a user (step S10). Then, the similar line selection unit 13 selects a model sentence similar to the input danger prediction activity sentence from the model sentence database 12 (step S11). Here, the similar line selection unit 13 selects a model sentence similar to the input danger prediction activity sentence based on, for example, the similarity between the work procedure items of the danger prediction activity sentence and the work procedure items of the model sentence. The similar line selection unit 13 calculates the similarity by comparing these two character strings using Word2vec or the like.

[0024] The item similarity calculation unit 14 calculates the similarity between the risk factor items and / or countermeasure items of the input hazard prediction activity sentence and the model sentence (step S12). Here, the item similarity calculation unit 14 calculates the similarity between the risk factor items of the hazard prediction activity sentence and the risk factor items of the model sentence using Word2vec or the like, and calculates the similarity between the countermeasure items of the hazard prediction activity sentence and the countermeasure items of the model sentence. The display target extraction unit 15 extracts combinations to be displayed from a plurality of combinations of danger prediction activity sentences and model sentences based on the similarity of each sentence (step S13).

[0025] Next, the sentence similarity calculation unit 16 calculates the overall similarity between the input hazard prediction activity sentence and the model sentence (step S14). Here, the sentence similarity calculation unit 16 calculates the similarity of the sentences, for example, by averaging the similarity of the work procedure items, the similarity of the risk factor items, and the similarity of the countermeasure items for each line. Finally, the comparison result display unit 17 displays the risk factor items and / or countermeasure items of the input hazard prediction activity sentence, displaying the background with a density according to the similarity of each item (step S15), and then the hazard prediction training support process shown in Figure 3 is terminated.

[0026] FIG. 4 is a diagram showing an example of a sentence input by a user. The danger prediction activity document 2 is composed of work procedure items 21, risk factor items 22, and countermeasure items 23.

[0027] Line 2a lists "Work preparation" as work procedure item 21. Risk factor item 22 lists "Taking a shortcut during preparation resulted in tripping over the actual machine and falling." Countermeasure item 23 lists "When preparing, move along the safe passage."

[0028] Line 2b lists "Work preparation" as work procedure item 21. Risk factor item 22 lists "Stumbling and falling on stairs when moving inside the building or going up or down stairs." Countermeasure item 23 lists "When going up or down stairs, divide tools into small pieces and have two people carry them so that you can easily see your feet, keep one hand free, and use the handrail."

[0029] In line 2c, "Restoring pipe bracing" is listed as work procedure item 21. "When untying the bracing wire, the buttocks of the hands suddenly come off with force and hit the face" is listed as risk factor item 22. "When untying the wire, wear protective glasses and gloves" is listed as countermeasure item 23.

[0030] Line 2d lists "Inspect the appearance of the piping" as work procedure item 21. Risk factor item 22 states, "When checking the markings on the hoisted piping, you may be distracted by what is above and trip over the piping at your feet, causing you to fall." Countermeasure item 23 states, "When inspecting the piping above, carefully check not only what is above but also your feet and surroundings."

[0031] Line 2e lists "pipe maintenance" as work procedure item 21. Risk factor item 22 lists "When polishing the groove with Gr, the air hose suddenly comes off and the Gr falls into the pipe." Countermeasure item 23 lists "When polishing the inside of the groove, attach a string to the Gr to prevent it from falling."

[0032] Line 2f lists "painting" as work procedure item 21. Risk factor item 22 states, "Because the work area was hot and the temperature was high, you sweat profusely while painting, and collapse from heatstroke." Countermeasure item 23 states, "Take appropriate breaks, replenish fluids and salt to prevent heatstroke."

[0033] FIG. 5 is a diagram showing an example of model sentence 3. The model text 3 includes a work procedure item 31 , a risk factor item 32 , and a countermeasure item 33 .

[0034] Line 3a lists "Work preparation" as work procedure item 31. Risk factor item 32 lists "Taking a shortcut during preparation resulted in tripping over the actual machine and falling." Countermeasure item 33 lists "When preparing, move along the safe passage."

[0035] In line 3b, "Work preparation" is listed as work procedure item 31. Risk factor item 32 states, "When carrying a load up and down stairs, you may hurt your back by carrying too much and doing something reckless." Countermeasure item 33 states, "Use an electrician's bag and move carefully, being careful of your step."

[0036] Line 3c lists "Re-braking pipes" as work procedure item 31. Risk factor item 32 lists "When handling wire, the rough wire may catch on fingers and cause cuts." Countermeasure item 33 lists "Always wear protective gloves when using wire."

[0037] Line 3d lists "pipe maintenance" as work procedure item 31. Risk factor item 32 lists "When maintaining pipes near an opening, you may become so engrossed in your work that you forget the opening exists, and accidentally fall while walking sideways." Countermeasure item 33 lists "When working near an opening, put on your safety belt first."

[0038] Line 3e states "External and internal inspection of pipes" as work procedure item 31. Risk factor item 32 states "When checking the markings on the pipes, the place is narrow, so when changing position, the body hits the surrounding pipes." Countermeasure item 33 states "When in a narrow place, carefully check the surrounding situation."

[0039] In line 3f, "painting" is listed as work procedure item 31. Risk factor item 32 states, "When painting the back of the pipe, the markings were painted over without checking." Countermeasure item 33 states, "Use masking tape to protect the markings."

[0040] FIG. 6 is a diagram illustrating the similarity calculation process. The upper part of Fig. 6 displays line 2a of the input hazard prediction activity text 2. The lower part of Fig. 6 displays model text 3. The similar line selection unit 13 calculates the similarity between the work procedure item 21 in line 2a and each of the work procedure items 31 in lines 3a to 3f. Each arrow indicates the operation of comparing line 2a with lines 3a to 3f.

[0041] FIG. 7 is a diagram illustrating the similarity calculation process. The upper part of Fig. 7 displays line 2a of the input hazard prediction activity text 2. The lower part of Fig. 7 displays model text 3, with only the work procedure items 31 in lines 3a and 3b displayed in white. The similar line selection unit 13 compares the similarity between the work procedure item 21 in line 2a and each of the work procedure items 31 in lines 3a to 3f, and selects lines 3a and 3b with the greatest similarity as the similar lines.

[0042] FIG. 8 is a diagram illustrating the similarity calculation process. The upper part of Fig. 8 displays row 2a of the entered hazard prediction activity sentence 2. The lower part of Fig. 8 displays model sentence 3, with work procedure items 31 and risk factor items 32 in rows 3a and 3b displayed in white. The item similarity calculation unit 14 calculates the similarity between the risk factor item 22 in row 2a and the risk factor items 32 in rows 3a and 3b, and selects the risk factor item 32 in row 3a with the greatest similarity as the similar item.

[0043] FIG. 9 is a diagram illustrating the similarity calculation process. The upper part of Fig. 9 displays row 2a of the input risk prediction activity sentence 2. The lower part of Fig. 9 displays model sentence 3, with work procedure item 31, risk factor item 32, and countermeasure item 33 in row 3a displayed in white. The item similarity calculation unit 14 calculates the similarity between countermeasure item 23 in row 2a and countermeasure item 33 in row 3a. The display target extraction unit 15 also selects row 3a as the display target.

[0044] 10A to 10C are diagrams showing similarity display screens. 10A displays lines 2a and 2b of the entered danger prediction activity sentence 2. Directly below line 2a, line 3a, which is similar to line 2a, of model sentence 3 is displayed. Directly below line 2b, line 3b, which is similar to line 2b, of model sentence 3 is displayed. In addition, the work procedure, risk factor, and countermeasure columns display the contents as text, and the similarity is displayed as a percentage. Each cell in the table is displayed with a hatching density that corresponds to the similarity. Specifically, since the similarity between the countermeasure in row 3b and the countermeasure in row 2b is 63.1%, the cell for the countermeasure in row 3b is displayed with hatching that corresponds to the similarity with row 2b. This allows the user to easily recognize the differences between the hazard prediction activity text 2 that they entered and the model text 3. However, the present invention is not limited to this, and the comparison result display unit may display the similarity using any of the table background, character color, font, and character thickness, and is not limited thereto.

[0045] 10B displays lines 2c and 2d of the entered danger prediction activity sentence 2. Directly below line 2c, line 3c, which is similar to line 2c, of model sentence 3 is displayed. Directly below line 2d, line 3d, which is similar to line 2d, of model sentence 3 is displayed. In addition, the work procedure, risk factor, and countermeasure columns display the contents as text, and the similarity is displayed as a percentage. Each cell in the table is displayed with a hatching density that corresponds to the similarity. Specifically, the similarity between the risk factor in row 3c and the risk factor in row 2c is 55.8%, so the cell for the risk factor in row 3c is displayed with hatching that corresponds to the similarity with row 2c.

[0046] Similarly, the cells for risk factors and countermeasures in row 3d are displayed with hatching that corresponds to the degree of similarity with row 2d. This allows the user to easily recognize the differences between the hazard prediction activity sentence 2 that they entered and the model sentence 3.

[0047] Figure 10C displays lines 2e and 2f of the entered danger prediction activity sentence 2. Directly below line 2e, lines 3d and 3e of model sentence 3, which are similar to line 2e, are displayed. Directly below line 2f, line 3f of model sentence 3, which is similar to line 2f, is displayed. In the work procedure, risk factor, and countermeasure columns, the content is displayed as text, and the similarity is displayed as a percentage. Each cell in the table is displayed with a hatching density that corresponds to the similarity.

[0048] Specifically, the similarity between the risk factor in row 3d and the risk factor in row 2e is 28.9%, so the cell for the risk factor in row 3d is displayed with hatching that corresponds to the similarity with row 2e.The similarity between the countermeasure in row 3d and the countermeasure in row 2e is 24.8%, so the cell for the countermeasure in row 3d is displayed with hatching that corresponds to the similarity with row 2e. Since the similarity between the work procedure in row 3e and the work procedure in row 2e is 60.8%, the cell of the work procedure in row 3e is displayed with hatching corresponding to the similarity.

[0049] Similarly, the cells for risk factors and countermeasures in row 3f are displayed with hatching that corresponds to the degree of similarity with row 2f. This allows the user to easily recognize the differences between the hazard prediction activity sentence 2 that they entered and the model sentence 3.

[0050] According to this embodiment, a support device for hazard prediction training can be provided that does not undermine the essence of hazard prediction activities, which is for each individual to consider danger, and that makes it easy to obtain educational materials. By collecting model hazard prediction sentences, registering them in the system, and comparing them with one's own hazard prediction sentences, necessary items and countermeasures can be confirmed for content with high similarity. On the other hand, awareness can be provided for content with low similarity. This makes it possible to quantify the level of hazard prediction training and improve the quality of hazard prediction training.

[0051] [1] a model sentence database (12) storing model sentences for risk prediction activities including work procedure items, risk factor items, and countermeasure items; an input unit (11) that accepts input of a danger prediction activity sentence by a user; a comparison result display unit (17) that displays risk factor items and / or countermeasure items of the danger prediction activity sentence input to the input unit (11), as well as model sentences similar to the danger prediction activity sentence and the degree of similarity with the danger prediction activity sentence in a distinguishable manner; A danger prediction training support device (1) comprising:

[0052] In this way, the danger prediction training support device can support danger prediction training in which each person thinks about the dangers that may occur during work.

[0053] [2] The comparison result display unit (17) displays the risk factor items and / or countermeasure items of the danger prediction activity sentence input to the input unit (11) as well as the similarity between the model sentence and the danger prediction activity sentence using any of the table background, character color, font, and character thickness. 2. The risk prediction training support device (1) according to claim 1.

[0054] This allows the danger prediction training support device to visualize the similarity of each item in an identifiable manner.

[0055] [3] The comparison result display unit (17) displays the work procedure items, risk factor items, and countermeasure items of the danger prediction activity text inputted to the input unit (11) in one row of a table. 2. The risk prediction training support device (1) according to claim 1.

[0056] This allows the hazard prediction training support device to segment and visualize each item in the hazard prediction activity sentence.

[0057] [4] The comparison result display unit (17) displays a model sentence similar to the danger prediction activity sentence input to the input unit (11) in one of the rows of the table, and displays the danger prediction activity sentence in the row adjacent to the row. 2. The risk prediction training support device (1) according to claim 1.

[0058] This allows the hazard prediction training support device to segment the hazard prediction activity sentence and model sentences similar to this hazard prediction activity sentence into each item and visualize them.

[0059] [5] a similar line selection unit (13) that selects model sentences similar to the danger prediction activity sentences input to the input unit (11) from the model sentence database (12); an item similarity calculation unit (14) that calculates the similarity between the risk factor items and / or countermeasure items of the risk prediction activity sentence input to the input unit (11) and the model sentence selected by the similar line selection unit (13); a sentence similarity calculation unit (16) that calculates the similarity of the model sentence that is similar to the danger prediction activity sentence based on the similarity determined by the item similarity calculation unit (14); a display target extraction unit (15) that extracts model sentences based on the similarity calculated by the sentence similarity calculation unit (16); The risk prediction training support device (1) according to claim 1, further comprising:

[0060] This allows the hazard prediction training support device to appropriately compare the hazard prediction activity sentence with the model sentence and calculate the similarity between them.

[0061] [6] The similar line selection unit (13) selects, from the model sentence database (12), a model sentence having a work procedure item similar to a work procedure item of the danger prediction activity sentence input to the input unit (11). 6. The risk prediction training support device (1) according to claim 5.

[0062] This allows the danger prediction training support device to appropriately select model sentences that are similar to the danger prediction activity sentences.

[0063] [7] The display target extraction unit (15) further includes a threshold setting unit (18) that sets a threshold value of the similarity when extracting combinations to be displayed. 6. The risk prediction training support device (1) according to claim 5.

[0064] This allows the risk prediction training support device to extract an appropriate model document for comparison as a display target.

[0065] [8] An input unit (11) receives an input of a danger prediction activity sentence from a user; A step in which a comparison result display unit (17) identifiably displays the risk factor items and / or countermeasure items of the danger prediction activity sentence input to the input unit (11), as well as model sentences similar to the danger prediction activity sentence and the degree of similarity with the danger prediction activity sentence; A hazard prediction training support method comprising:

[0066] This will support risk prediction training, where each person thinks about the risks involved in their work.

[0067] <<Variation>> The present invention is not limited to the above-described embodiments and includes various modifications. For example, the above-described embodiments have been described in detail to clearly explain the present invention, and are not necessarily limited to those including all of the described configurations. It is possible to replace part of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add the configuration of another embodiment to the configuration of one embodiment. Furthermore, it is also possible to add, delete, or replace part of the configuration of each embodiment with other configurations.

[0068] The above-described configurations, functions, processing units, processing means, etc. may be realized in part or in whole by hardware such as an integrated circuit. The above-described configurations, functions, etc. may be realized by software by a processor interpreting and executing a program that realizes each function. Information such as the programs, tables, and files that realize each function can be stored in a storage device such as a memory, a hard disk, or an SSD (Solid State Drive), or on a storage medium such as a flash memory card or a DVD (Digital Versatile Disk).

[0069] In each embodiment, the control lines and information lines shown are those that are considered necessary for the explanation, and not all control lines and information lines in the product are necessarily shown. In reality, it can be considered that almost all components are interconnected. [Explanation of symbols]

[0070] 1. Hazard prediction training support device 11 Input section 12 Model Sentence Database 13 Similar bank selection section 14 Item similarity calculation part 15 Display target extraction section 16 Sentence similarity calculation part 17 Comparison result display area 18 Threshold setting section 91 CPU 92 ROM 93 RAM 94 Memory section 95 Display section 96 Input section 97 Communication Interface 98 Bus 2. Hazard prediction activity document 21 Work procedure items 22 Risk Factor Items 23 Countermeasures 2a~2f rows 3 Model sentences 31 Work procedure items 32 Risk Factor Items 33 Countermeasures 3a~3f rows

Claims

1. a model sentence database storing model sentences for risk prediction activities including work procedure items, risk factor items, and countermeasure items; an input unit that accepts input of a danger prediction activity sentence by a user; A comparison result display unit that displays the work procedure items and / or risk factor items and / or countermeasure items of the danger prediction activity sentence input to the input unit, as well as model sentences similar to the danger prediction activity sentence and the degree of similarity with the danger prediction activity sentence in a distinguishable manner; A danger prediction training support device comprising:

2. The comparison result display unit displays the risk factor items and / or countermeasure items of the danger prediction activity sentence inputted to the input unit, as well as the similarity between the model sentence and the danger prediction activity sentence, using any of the table background, character color, font, and character thickness.

2. The risk prediction training support device according to claim 1.

3. The comparison result display unit displays the work procedure items, risk factor items, and countermeasure items of the danger prediction activity text inputted to the input unit in one row of a table.

2. The risk prediction training support device according to claim 1.

4. The comparison result display unit displays a model sentence similar to the danger prediction activity sentence input to the input unit in one of the rows of the table, and displays the danger prediction activity sentence in a row adjacent to the row.

2. The risk prediction training support device according to claim 1.

5. a similar line selection unit that selects model sentences similar to the danger prediction activity sentences inputted to the input unit from the model sentence database; an item similarity calculation unit that calculates the similarity between the risk factor items and / or countermeasure items of the risk prediction activity sentence inputted to the input unit and the model sentence selected by the similar line selection unit; A sentence similarity calculation unit that calculates the similarity of the model sentence that is similar to the risk prediction activity sentence based on the similarity determined by the item similarity calculation unit; a display target extraction unit that extracts model sentences based on the similarity calculated by the sentence similarity calculation unit; 2. The risk prediction training support device according to claim 1, further comprising:

6. The similar line selection unit selects, from the model sentence database, a model sentence having a work procedure item similar to a work procedure item of the danger prediction activity sentence input to the input unit.

6. The risk prediction training support device according to claim 5.

7. a threshold setting unit that sets a threshold value for the similarity when the display target extraction unit extracts combinations to be displayed; 6. The risk prediction training support device according to claim 5.

8. An input unit receives an input of a danger prediction activity sentence from a user; A step in which a comparison result display unit identifiably displays the work procedure items and / or risk factor items and / or countermeasure items of the danger prediction activity sentence inputted to the input unit, as well as model sentences similar to the danger prediction activity sentence and the degree of similarity with the danger prediction activity sentence; A hazard prediction training support method comprising:

Images