Method and system for dynamically determining a risk index of a geographical area

Abstract

Method ( 100 ) implemented through a distributed computer system (2 ) to dynamically determine at least one travel or stay risk index in at least one geographical area, comprising the steps of : - defining ( 101 ) a plurality of risk categories, each risk category corresponding to a type of risk to which a person may be exposed when travelling or staying in said geographical area; - defining ( 102 ) for each risk category a plurality of keywords semantically related to the risk category and storing said plurality of keywords in a data structure; - for each risk category, extracting ( 103 ) first risk data by consulting informative websites and social media and carrying out a computer sentiment analysis of first textual contents published in, or associated with, a first relatively long time period on said informative websites and social media, where the first textual contents concern and / or mention said geographical area and are selected using the keywords semantically related to the risk category; - determining ( 104 ) a basic risk index associated with the geographical area starting from the first risk data; - receiving or generating a ( 106) request to determine a. n instantaneous risk index associated with the geographical area; - for each risk category, extracting ( 107 ) second risk data by consulting informative websites and social media and carrying out a computer sentiment analysis of second textual contents published in, or associated with, a second relatively short time period on said informative websites and social media, where the second textual contents concern and / or mention said geographical area and are selected using the keywords semantically related to the risk category; - determining ( 108 ) an instantaneous risk index associated with the geographical area starting from the second risk data; - comparing ( 109 ) the instantaneous risk index and the basic risk index; generate and send ( 110 ) a notification and / or control signal if from the comparing step it is found that the instantaneous risk index is higher than the basic risk index

Description

Method and system for dynamically determining a risk index of a geographical areaDESCRIPTIONTechnical field of the invention

[0001] The present description refers to the technical field of data processing methods and systems and relates, in particular, to a method and a system for dynamically determining a risk index of a geographical area. For the purposes of this description, dynamic determination (or "dynamic assessment" ) of a risk index is specifically understood as the ability to assess whether, in a given geographical area and in a given time period, a risk index representative of one or more risk categories to which a subject may be exposed is greater than, less than, or equal to a risk index to which the same subject would normally be exposed in the same geographical area but in other time periods.Technological background of the invention

[0002] As is well known, a geographical area can expose one or more people to various categories of risk, such as, for example and not limited to:• socio-political risks: riots, wars, unrest, attacks, acts of terrorism;• crime and petty crime: theft, robbery, shooting, mugging, pickpocketing;• health: epidemics, pandemics, other risks related to people ' s health;• climate and environmental / geological risks: chemical, biological, physical, landslide, fire, earthquake, flood, volcanic eruption risks.

[0003] Historically, the assessment of these risks is made over a relatively long period of time based on statistical data collected by various oraganizations, for example, institutional or governmental (such as in Italy: Ministry of the Interior; Ministry of Foreign Affairs; ISTAT). The updating times ofstatistical data are typically of the order of a year and reflect the fact that the massive collection of data on the basis of which the statistics are processed requires both long times, of the order of months, and numerous hours of human work.

[0004] However, given the rapid dynamics that can often characterise the evolution of risks, a long-term statistical assessment of the risks to which a geographical area is exposed and which is subject to a relatively slow update, is currently considered unsatisfactory in relation to various needs. In fact, in some cases, there is a need to have much more precise information on the risk of travelling or staying in a specific geographical area available in a short time.

[0005] For example, this information may be useful to one or more individuals and / or entities such as:• law enforcement agencies, civil protection, institutions, event organisation bodies;• individuals who travel to a given geographical area, other than the one where they usually reside, and can be made aware of risk conditions against which they can protect themselves;• individuals (both natural and legal persons) residing in a certain geographical area, who may not be aware of the existence, at particular times, of risk conditions against which they should protect themselves;• travel agencies, which, depending on the destinations of their customers, can inform them of the onset of higher than normal risk conditions, and provide for risk mitigation policies;• employers, interested in mitigating the risks to which their staff may be exposed when travelling;• Insurance companies, which can offer people, either directly or through travel agencies or employers, policies dedicated to covering specific risks that arise at certain times and in certain areas;• automated systems for generating alarms and / or alerts or automated active systems for controlling and / or mitigating risk, for example traffic control systems, dams, barriers, optical oracoustic signalling systems, etc.

[0006] There is therefore a need to be able to quickly and massively collect useful information to dynamically determine a risk index for a geographical area.

[0007] The purpose of this invention is to provide a method and a system for dynamically determining at least one risk index that is able to satisfy the aforementioned need.

[0008] This and other purposes are achieved by a method for dynamically determining a risk index as defined in claim 1 in its most general form, and in the claims dependent on it in some particular embodiments. These purposes are also achieved through a distributed computer system as defined in claim 12.Brief description of the drawings

[0009] The invention will be better understood from the following detailed description of its embodiments, given by way of example and therefore in no way limiting in relation to the accompanying drawings, in which:- figure 1 shows a simplified logical block diagram of a distributed computer system configured to dynamically determine at least one travel or stay risk index in at least one geographical area according to a non-limiting embodiment of the present invention;- figure 2 shows the result of an example of clustering that can be used in the aforementioned dynamic determination;- figure 3 shows the result of a further example of clustering that can be used in the aforementioned dynamic determination; - figure 4 shows a flow chart of a non-limiting embodiment of a method, executable by the distributed computer system of figure 1, to dynamically determine at least one travel or stay risk index in at least one geographical area.

[0010] In the attached figures, the same or similar elements may be indicated by the same reference signs.

[0011] It is specified here that elements of different embodiments can be combined together to provide further embodiments withoutlimits while respecting the technical concept of the invention, as the average person skilled in the art understands directly and unambiguously from what is described.

[0012] This description also refers to the technique known for its implementation, with regard to the detailed features not described, such as elements of minor importance usually used in the technique known in solutions of the same type.When an element is introduced, it is always understood that it can be "at least one" or "one or more".

[0013] When a list of elements or features is listed in this description, it is meant that the matter according to the invention "comprises" or alternatively "is composed of" such elements. When listing features within the same sentence or bulleted list, one or more of the individual features may be included in the invention without connection to the other features in the list.

[0014] Two or more of the parts (elements, devices, systems) described below can be freely associated and considered as kits of parts according to the invention.

[0015] Figure 1 shows an illustrative and non-limiting embodiment of a distributed computer system 1. The distributed computer system 1 comprises a cloud computing infrastructure 2, at least one data supply system Dp1, Dp2, Dp3, DpN and at least one mobile user terminal 3. Cloud computing infrastructure 2 means an infrastructure comprising hardware and software resources for data processing, data storage and data communication via at least one telecommunications network 4, 5.

[0016] The at least one telecommunications network 4, 5 can be a wireless network or a wired network or a hybrid network that is partly wireless and partly wired. For example, at least part of the telecommunications network 4, 5 is a cellular radio network 4, for example a 4G, LTE or 5G network. Said cellular radio network 4 allows the mobile user terminal 3 to be operationally connected with the cloud computing infrastructure 2.

[0017] The cloud computing infrastructure 2 comprises a backend server 20, responsible for data processing, and at least onedatabase 21 accessible from the backend server 20. The cloud computing infrastructure 2 is configured to determine, for at least one geographical area, a basic risk index and an instantaneous risk index and to compare said risk indices with each other. Preferably, said risk indices are numerical risk indices. In an alternative embodiment, said risk indices are alphanumeric indices to which respective numerical values are associated.

[0018] According to an advantageous embodiment, the cloud computing infrastructure 2 is configured to determine a respective basic risk index for each of a plurality of geographical areas and to store the basic risk indices in a GIS (Geographic Information System) database 22 included in the cloud computing infrastructure 2.

[0019] According to an advantageous embodiment, the risk indices (base and instantaneous) are determined by taking into consideration a plurality of risk categories to which a person may be exposed during a trip or a stay in at least one geographical area.

[0020] These risk categories are, for example, each representative of a risk belonging to the types of risks listed below, by way of example, non-limiting and non-exhaustive:• socio-political risks: riots, wars, unrest, attacks, acts of terrorism;• crime and petty crime: theft, robbery, shooting, mugging, pickpocketing;• health: epidemics, pandemics, other risks related to people ' s health;• climate and environmental / geological risks: chemical, biological, physical, landslide, fire, earthquake, flood, volcanic eruption risks.

[0021] The cloud computing infrastructure 2 comprises at least one web API 25, 26. According to an advantageous embodiment, the at least one web API 25, 26 comprises a first web API 25 configured to operatively connect the cloud computing infrastructure 2 tothe at least one data supply system Dp1, Dp2, Dp3, DpN and a second web API 26 configured to operatively connect the cloud computing infrastructure 2 to the user mobile terminal 3. It goes without saying that in a real implementation the second web API 26 is configured to connect the cloud computing infrastructure 2 to a multitude of mobile user terminals 3.

[0022] The mobile user terminal 3 is, for example, a mobile personal communication device, such as a smartphone, a tablet PC, a wearable communication device, such as a smartwatch. Alternatively, the mobile user terminal 3 can also be a personal computer. In a manner known per se and for this reason not further described, the mobile user terminal 3 comprises at least one processing unit, such as a microprocessor, a processor or a SoC (System on Chip) and at least one memory unit connected to, or integrated in, the processing unit. The mobile user terminal 3 may comprise a display, for example a touchscreen display. The mobile user terminal 3 comprises at least one communication interface, for example a wireless communication interface, in order to be operatively connected to the cloud computing infrastructure 2, for example via a telecommunications network 4. For example, the communication interface is a long-range radio communication interface and, for example, comprises a radio-cellular modem, such as a 4G, LTE or 5G modem.

[0023] According to an advantageous embodiment, an application program 30 (or " APP" ) is installed onboard the user mobile terminal 3 which allows the user mobile terminal 3 to be operatively connected to the cloud computing infrastructure 2, for example via the web API 26.

[0024] The one or more data supply systems Dp1,...,DpN are configured to acquire, from various data sources, risk data divided by time periods and geographical areas and preferably also divided by risk categories. These data supply systems Dp1,...,DpN can be managed, for example, by data providers or by the same organisation that manages the cloud computing infrastructure 2.

[0025] In recent times, new techniques have emerged for the mass acquisition of data, which are based on the collection of datapresent on informative websites (news agencies, online newspapers, etc. ) and on social media (e. g. Facebook, Instagram, X) and on the automated extraction of relevant information content (so-called "web scraping" techniques). Web scraping is complemented by the so-called "sentiment analysis" technique, which searches for keywords in published contributions that appear to be related to the type of information to be collected. In the specific case of determining risks, one can, for example, search for keywords related to concepts of fear, concern, alertness, etc.

[0026] Compared to traditional risk assessment, the use of web scraping techniques allows a risk assessment that can be updated practically in real time, since the time required to publish new contents on websites and social networks and to collect and process the relevant updated data is in the order of hours rather than months, as is the case with traditional techniques.

[0027] The present invention performs a dynamic assessment of the risks of staying or travelling in a geographical area that exploits techniques based on web scraping, and optionally also traditional techniques, and advantageously includes the following steps:- a set of risks of interest is defined, belonging to one or more of the risk categories such as those listed above, and / or to other risk categories; the risk categories will also be referred to below as "topics";- a tessellation of a territory of interest (for example, Italy) is defined, in such a way that the entire territory of interest is divided into geographical areas considered homogeneous from the point of view of the risks of interest; the tessellation is preferably periodically reviewed according to the results of the following steps;- a risk analysis is launched on each tile (or "geographical area" ) to determine the so-called "basic risk index", i. e., for each of the defined topics, the average risk resulting over a relatively long period (e. g. 6 to 12 months) in each geographical area of the tessellation is assessed and / or estimated and / or calculated; data sources fed by web scraping and optionally also traditionalsources are used; the results are advantageously stored in the GIS database 22, i. e. in a database set up to be queried on a geographical basis; the basic risk index is updated by repeating the analysis at regular intervals (e. g. every 3 months);- for example, when a given subject (for example, a person travelling) reaches (or expects to reach in a short time) a specific geographical area, the instantaneous determination of the risk situation is launched in relation to the topics defined for that geographical area; use is made of data sources fed by web scraping (as well as optionally traditional sources), where, however, in particular those fed by web scraping will be able to provide updated results that may highlight situations of exposure to risk other than those of the basic risk for that same geographical area;- the result of the analysis is then made available to the interested party by the appropriate means (e. g. on the smartphone mobile terminal user 3 when the interested party is a natural person, or through IT systems when the interested party is a travel agency, or an insurance company, an employer, law enforcement agencies, automated risk control systems, warning or alarm systems, etc. ).

[0028] The tessellation can be modified over time based on the results of the above analyses: geographical areas that recurrently show the same levels of risk exposure can be combined into a single tile, while areas where a strong variation in risk is identified between contiguous tiles can be divided into several tiles. This update follows the periodicity of the update of the basic risk determination (e. g. quarterly). The geographical areas defined by the tessellation can be, for example, neighbourhoods, towns, cities, districts, regions, groups of regions, nations, groups of nations.

[0029] The list of topics considered of interest may also vary over time, for example due to the availability of new types of risk identified and detectable with traditional techniques and / or web scraping.

[0030] The logical centre of the system is the cloud computing infrastructure 2 which includes the backend server 20, at least one database 21, and web APIs (or interfaces) 25, 26 for data exchange with external subsystems.

[0031] Among the external subsystems, there may be one or more data supply systems Dp1, ..., Dp2 (for example managed by data providers, and for this reason henceforth also called data providers) that provide information for risk assessment, both in relation to the basic risk index and in relation to the instantaneous risk index. The data providers may be suppliers of services existing on the market, even conceptually outside the perimeter of the system that is the subject of the present invention but in any case essential for the operation of the system, similarly to other external systems that will be discussed below, such as GNSS localisation systems (e. g., GPS) and cellular networks.

[0032] Among the data providers, a distinction can be made between traditional data providers and so-called web scrapers, which search for data on the web. In both cases, the Web API 25 may comprise at least two web APIs 25 having slightly different implementations.

[0033] In the case of traditional data providers, web APIs are typically exposed where the geographical query refers to a specific geographical area, for example an administrative area (e. g. a certain municipality, or a certain province, or even a certain region), a period of time (typically month or year), and the response that is provided is the data for that period in that geographical area, such as, for example, statistics on reports of theft, robbery, etc. This is done for each geographical area envisaged by the tessellation.

[0034] In the case of web scrapers, the web APIs 25 exposed are designed slightly differently. To determine the basic risk index, the geographical information of each tile, the required time span (e. g. last 6 or 12 months), and the topics for which the result of the sentiment analysis is required are sent to the data provider via the API 25. The data provider responds by providing a sentimentresult evaluated over the requested time period via the web API 25, for example a score (typically between 0 and 100) that represents how low (result close to zero) or high (result close to 100) that risk appears in the media and social media analysed. For the determination of the instantaneous risk, the process is the same, but the request is made by providing the specific location of an interested party (e. g., latitude, longitude, radius of uncertainty) and a current timestamp. The response provided by the data provider is a sentiment result evaluated in the last few hours rather than over a period of 6 to 12 months.

[0035] Among the data providers there may be one (or more than one) that provides the geolocation service on a mobile network cell basis. Under certain conditions, which will be better specified below, it is possible that the position of the person involved in the instantaneous risk assessment is known only in terms of the unique identifier of the mobile network cell where the mobile user terminal 3 is located (possibly in conjunction with the identifiers of neighbouring cells and relative signal levels). This (or these) data provider (s) provide the geographical position (estimated latitude and longitude) on the basis of the information relating to the cell where the mobile terminal user 3 of the person concerned is registered and possibly on the basis of information relating to neighbouring cells. In this case, the API allows the data provider to be informed of the information relating to one or more mobile network cells, and provides in response the estimated geographical coordinates for that signal scenario received from the mobile user terminal 3.

[0036] Similarly, for reasons that will be explained below, one or more data providers may offer the geocoding service, i. e. the possibility, given a postal address (possibly limited to the name of a locality), of receiving the estimated geographical coordinates for that address. In this case, the API allows the communication of information relating to an address to the data provider, and provides the relative geographical coordinates in response.

[0037] In all cases (both on the basis of data obtained fromtraditional data providers and from web scraping, as well as in the case of calculating the basic risk index or the instantaneous risk index) the backend server 20, with traditional techniques, advantageously performs a clustering of the results for each topic, i. e. it groups the results into classes useful for assessing a significant variation in risk between the basic risk index and the instantaneous risk index. Figure 2 represents a hypothetical data clustering for the basic risk index on three topics, respectively named in the figure "Topic 1", "Topic 2" and "Topic 3".

[0038] In this schematic, each small circle represents a tile into which the entire geographical area of interest is divided. For each tile and for each topic, the so-called "sentiment score" is calculated, which is preferably a normalised score between zero and one and represents the greater or lesser risk associated with that topic, and which is preferably calculated as a weighted average of the results obtained from the various data sources and / or for the various keywords that are associated with that sentiment. The circles enclosed in the oval gl represent the "low risk" cluster, those enclosed in the oval g2 the "medium risk" cluster, and those enclosed in the oval g3 the "high risk" cluster. The circles with a superimposed cross represent a specific location on which attention is focused for the subsequent determination of the instantaneous risk index. In this example, the location was classified as low risk for Topic 1, high risk for Topic 2, and medium risk for Topic 3.

[0039] In the case of determining the instantaneous risk index, the same calculations are carried out, but the result for the location in question can be modified, due to the effect of risk situations present at that time, as represented in the example in Figure 3.

[0040] In this example, it is observed that the level of risk for the location in question does not change with respect to Topic 1 and 3, and instead decreases (from high and medium) for Topic 2. If the opposite had happened, i. e. if the assessment for Topic 2 had gone from medium to high, a situation of increased risk wouldhave been identified with respect to Topic 2.

[0041] To perform the processing described above, the backend server 20 advantageously comprises databases 21, 22, 23 diversified according to the nature of the data being processed. Specifically, the information relating to the tessellation and the results of the basic risk assessment are stored in the GIS geographic database 22. Typically, a GIS layer is dedicated to tessellation (and related updates), and each layer is associated with a topic with the relative sentiment score values for each tile.

[0042] The information relating to the instantaneous risk assessments is stored and historicised in a database 23 where the data relating to the users is pseudonymised. This reflects the fact that the current regulations on the processing of personal data consider location data as data which, although not considered "sensitive", must be treated with particular security measures.

[0043] The personal data of the users and the pseudonymisation keys that allow tracing from the pseudonym to the corresponding user are stored in a dedicated database 24 subject to specific security measures and access restrictions. The separation between localization information and pseudonymisation keys constitutes a security measure that makes it difficult for an unauthorised person to link the localization information to a uniquely identifiable person.

[0044] The determination of the basic risk index is launched at regular intervals (for example, every three months with an analysis depth of six months or one year). The determination of the instantaneous risk index is instead preferably launched on an event (" Trigger Event" ) which generally originates from the mobile user terminal 3 of the person interested in knowing the determination of the instantaneous risk index, on which the specific APP 30 is installed. The APP 30 is preferably configured to communicate a trigger event to the backend server 20 via a suitable API 30 in the following circumstances:1) when the GNSS receiver of the mobile user terminal 3 detectsthat the user is in a location other than the one where he / she usually resides;2) when, through information from the mobile network via the mobile network operator, it is detected that the user is in a location other than that where he / she usually resides; 3) when the user, through a dedicated user interface made available by the APP 30, declares that he / she has gone (or is going) to a location other than the one where he / she usually resides, or wishes the analysis of the instantaneous risk to be carried out for the location where he / she is located, even if this coincides with the location where he / she usually resides.

[0045] In case 1) the APP 30 installed on the mobile user terminal 3 sends the geographical coordinates detected by the GNSS receiver to the backend server 20. In case 2), if the geographical coordinates are not available (e. g. because the smartphone ' s GNSS receiver is disabled or faulty), the APP sends the backend server 20 the identifiers of the current mobile network cell and possibly those of the neighbouring cells and the relative signal levels where available; the backend server 20 will convert this data into the corresponding estimated geographical coordinates using a data provider that provides the geolocation service on a mobile network cell basis. In case 3, the user enters the address of the location of interest (or selects a "use current location" option) via the APP 30 user interface, which the backend server 20 receives and converts into the corresponding geographical coordinates through the data provider that provides the geocoding service.

[0046] Once the coordinates of the location for which the instantaneous risk assessment is to be carried out have been determined, the backend server 20 performs said assessment using the data providers Dpi,..., DpN as previously described. The value of the instantaneous risk index thus determined is compared with the current value of the basic risk stored in the GIS database 22, and the result of the comparison, with any variations in risk found on the various topics, generates a notification. For example, this notification is communicated to the user via thesame APP 30. This notification, for example, informs the user of an increase in the risk index. Alternatively or additionally, this notification is sent to an automatic risk control or mitigation system.

[0047] Advantageously, one or more web interfaces for various types of operators are conceptually part of the cloud computing infrastructure 2, but are implied in the general architecture, such as:- operators who administer the system, with permissions organised on several levels (for example, only higher-level operators can access the database 24 of the pseudonymisation keys and thus become aware of the identity of the users);- corporate users, such as travel agencies, insurance companies, and employers, who can request and / or monitor the risk of users connected to them, and therefore take any actions to mitigate the risks.

[0048] With reference to Figure 4, we will now describe a method 100 implemented through a distributed computer system, such as the distributed computer system 1 described so far or similar or equivalent systems, to dynamically determine at least one travel or stay risk index in at least one geographical area. Further features of method 100 that are evident from the above description for the distributed computer system 1 may not be repeated for the sake of brevity.

[0049] The method 100 comprises a step 101 of defining a plurality of risk categories, each risk category corresponding to a type of risk to which a subject may be exposed when travelling or staying in said geographical area. As already explained, these types of risk, or "topics", may include one or more of the types of risks listed below:- socio-political risks;- crime-related risks;- health-related risks;- climate risks and environmental / geological risks.

[0050] The method 100 comprises a step of defining 102 for eachrisk category a plurality of keywords (e. g. a list of keywords) semantically related to the risk category and storing said plurality of keywords in a data structure. This data structure can be stored in the cloud computing infrastructure 2 and / or in the data supply systems Dbl,..., DbN.

[0051] For each risk category, the method 100 includes a step of extracting 103 first risk data by consulting informative websites and social media and carrying out a computer sentiment analysis of first textual contents published in, or associated with, a first relatively long time period on said informative websites and social media, in which the first textual contents concern and / or mention said geographical area and are selected by means of the keywords semantically related to the risk category. The relatively long time period is a period of the order of months, for example equal to six or twelve months. Published in, or associated with, means that in some way the content pertains to the time period of interest, for example either because it has a publication date (explicit in the content or reported in metadata associated with the content) falling within the period of interest or because the text of the content mentions a date falling within the period of interest, etc. According to an advantageous embodiment, the sentiment analysis is carried out by means of at least one algorithm or web-scraping service.

[0052] According to an advantageous embodiment, the aforementioned step 103 of extracting first risk data is carried out by also querying one or more computer databases Db1,..., BdN in which information on risk levels associated with the risk categories and concerning said relatively long time period is stored for said geographical area. It should be noted that these computer databases represent data sources that can be consulted using the techniques that have previously been defined as traditional techniques.

[0053] According to an advantageous embodiment, the method 100 comprises, preferably in the step of extracting 103, an operation of defining a tessellation of a territory of interest, in such a way that the entire territory of interest is divided intogeographical areas considered homogeneous from the point of view of the risks of interest and in which said geographical area for which the method 100 dynamically determines at least one travel or stay risk index is one of said geographical areas. For further details on tessellation, see what has already been described for the distributed computer system 1.

[0054] The method 100 also includes a step of determining 104 a basic risk index associated with the geographical area starting from the first risk data. For example, in said step of determining 104 the basic risk index is calculated as a weighted average of a plurality of risk indices.

[0055] According to an advantageous embodiment, through the steps 101-104 described so far, the method 100 allows the determination of a respective basic risk index for each of a number of geographical areas. The basic risks and the information that allows the geographical areas to be identified are preferably stored in the GIS database 21. For each geographical area, it is possible to determine a number of basic risk indices, each associated with a respective risk category.

[0056] The method 100 includes a step of receiving or generating 106 a request to determine an instantaneous risk index associated with the geographical area, or generally with a geographical area of interest. For example, the determination request is received or generated by the backend server 20, or generally by the cloud computing infrastructure 2. Receiving means that an entity external to the backend server 20, for example the mobile user terminal 3, generates this request and sends it to the backend server 20. Generating means that the request is generated automatically or on the basis of an input from the backend server 20.

[0057] The method 100 for each risk category includes a step of extracting 107 second risk data by consulting informative websites and social media and carrying out a computer sentiment analysis of second textual contents published in, or associated with, a second relatively short time period on said informative websites and social media, in which the second textual contents concernand / or mention said geographical area and are selected by means of the keywords semantically related to the risk category. This step 107 is identical or similar to the step 102 of extracting the first risk data, with the only difference that it is carried out for a second shorter time period, preferably much shorter, than the first time period. The second time period is, for example, of the order of hours or tens of hours, and is, for example, equal to twelve or 24 hours. Also in this step 107 of extracting, the sentiment analysis is advantageously carried out by means of at least one web-scraping algorithm or service.

[0058] Advantageously, the step of extracting 107 second risk data is also carried out by further querying one or more computer databases Dbl,..., DbN in which information on risk levels associated with the risk categories and concerning said relatively short time period is stored for said geographical area. It should be noted that also in this case said computer databases represent data sources that can be consulted using the techniques that have been previously defined as traditional techniques.

[0059] According to an advantageous embodiment, the request for determination of the instantaneous risk index that is received or generated in step 106 comprises a timestamp that is used as a time reference to determine the second time period.

[0060] The method 100 also comprises a step of determining 108 an instantaneous risk index associated with the geographical area starting from the second risk data. For example, in this step of determining 108 the instantaneous risk index is calculated as a weighted average of a plurality of risk indices.

[0061] It should be noted that the step of receiving or generating 106 the request to determine the instantaneous risk index associated with the geographical area, or in general with a geographical area of interest, is the event that determines the start of steps 107 and 108, i. e. the trigger event already described above.

[0062] Conveniently, the method 100 comprises a step 105 of acquiring positioning and / or localization information of a mobileuser terminal 3 and in which the request for determining the instantaneous risk index associated with the geographical area comprises said positioning and / or localization information or information that makes it possible to identify that the mobile user terminal 3 is in said geographical area and that is obtained through said positioning and / or localization information. Conveniently, said request is sent from said mobile user terminal 3, preferably via the APP 30 installed on board said mobile user terminal 3. Advantageously, said determination request is received or generated as soon as it is detected that said mobile user terminal 3 is located in said geographical area.

[0063] The method 100 includes a step of comparing 109 the instantaneous risk index and the basic risk index with each other and a step of generating and sending 110 a notification and / or control signal if from the step of comparing 109 it is detected that the instantaneous risk index is higher than the basic risk index. For example, the notification and / or control signal is an alert and / or alarm signal and said signal is sent through a telecommunications network 4, for example to the mobile user terminal 3 and / or to a risk control and / or mitigation system and / or to an alert or alarm system.

[0064] For example, if, on the other hand, the comparison step 109 reveals that the instantaneous risk index is lower than or equal to the base risk, it may be decided not to take any action or to store this information in a database (e. g. for the purpose of monitoring the evolution of an area ' s risk index), e. g. in the GIS database 22, or to send a notification or control signal in any case.

[0065] Based on the above, it is therefore possible to understand how the proposed method and system allow the intended aims to be fully achieved with reference to the state of the art.

[0066] Without prejudice to the principle of the present invention, the forms of implementation and the details of realisation may be widely varied with respect to what has been described and illustrated purely by way of non-limiting example, without thereby departing from the scope of protection as definedin the appended claims.

Claims

CLAIMS1. A method (100) implemented by a distributed computer system (2) for dynamically determining at least one travel or stay risk index in at least one geographical area, comprising the steps of:- defining (101) a plurality of risk categories, each risk category corresponding to a type of risk to which a person may be exposed when travelling or staying in said geographical area;- defining (102) for each risk category a plurality of keywords semantically related to the risk category and storing said plurality of keywords in a data structure;- for each risk category, extracting (103) first risk data by consulting informative websites and social media and carrying out a computer sentiment analysis of first textual contents published in, or associated with, a first relatively long time period on said informative websites and social media, where the first textual contents concern and / or mention said geographical area and are selected using the keywords semantically correlated to the risk category;- determining (104) a basic risk index associated with the geographical area starting from the first risk data;- receiving or generating a (106) request to determine an instantaneous risk index associated with the geographical area;- for each risk category, extracting (107) second risk data by consulting informative websites and social media and carrying out a computer sentiment analysis of second textual contents published in, or associated with, a second relatively short time period on said informative websites and social media, where the second textual contents concern and / or mention said geographical area and are selected using the keywords semantically correlated to the risk category;- determining (108) an instantaneous risk index associated with the geographical area starting from the second risk data;- comparing (109) the instantaneous risk index and the basic risk index with each other;- generating and sending (110) a notification and / or controlsignal if from the comparing step it is found that the instantaneous risk index is higher than the basic risk index.

2. Method (100) according to claim 1, wherein said determination request comprises a timestamp which is used as a time reference to determine the second time period.

3. Method (100) according to claims 1 or 2, wherein said notification and / or control signal is an alert and / or alarm signal and wherein said signal is sent through a telecommunications network ( 4 ).

4. Method (100) according to any one of the preceding claims, comprising a step (105) of acquiring positioning and / or localization information of a mobile user terminal (3) and wherein the request for determining the instantaneous risk index associated with the geographical area comprises said positioning and / or localization information or information that makes it possible to identify that the mobile user terminal (3) is in said geographical area and that is obtained through said positioning and / or localization information.

5. Method (100) according to claim 4, wherein said request is sent by said mobile user terminal (3), preferably via an APP (30) installed on board said mobile user terminal (3).

6. Method (100) according to claims 4 or 5, wherein said request is sent or generated as soon as it is detected that said mobile user terminal (3) is located in said geographical area.

7. Method (100) according to any one of the preceding claims, wherein said first relatively long time period is of the order of months and said second relatively short time period is of the order of days or tens of hours.

8. Method (100) according to any one of the preceding claims, wherein:- said step (103) of extracting first risk data is carried out by also querying one or more computer databases in which, for said geographical area, information on risk levels associated with the risk categories and concerning said first relatively long time period is stored;- said step of extracting (107) second risk data is carried out by also querying one or more computer databases in which, for said geographical area, information on risk levels associated with the risk categories and concerning said second relatively short time period is stored.

9. Method (100) according to any one of the preceding claims, wherein said sentiment analysis is carried out by means of at least one web-scraping algorithm.

10. Method (100) according to any one of the preceding claims, wherein said instantaneous risk index and said basic risk index are numerical risk indices.

11. Method (100) according to any one of the preceding claims, comprising, preferably in said step of extracting (103), an operation of defining a tessellation of a territory of interest, in such a way that said territory of interest is divided into geographical areas deemed homogeneous from the point of view of the risks of interest and in which said geographical area for which the method (100) dynamically determines at least one travel or stay risk index is one of said geographical areas.

12. Distributed computer system (1) comprising a cloud computing infrastructure (2) configured to execute a method (100) according to any one of the preceding claims.

Images