FIELD OF THE INVENTION

The present invention relates to a computer-implemented event tracking and forecasting system and a method of using the system. More particularly, but not exclusively, the invention relates to a computer-implemented crime tracking and forecasting system and a method of using the system to obtain data on a possible future crime.

BACKGROUND TO THE INVENTION

Crime is a persistent problem all over the globe and has a negative impact on quality of life and economic growth. Law enforcement agencies are often overwhelmed and, more specifically, lack sufficient manpower to prevent a crime prior to it being committed by a perpetrator.

Conventional methods of detecting and forecasting possible crimes include surveillance conducted by officers of a law enforcement agency. It will be appreciated that a surveillance operation of the “stakeout” kind requires boots on the ground. Many law enforcement agencies simply lack the financial and human resources to effectively survey criminal suspects throughout a relevant jurisdiction. It would be preferable if a law enforcement agency could focus its financial and human resources on suspects and locations within a relevant jurisdiction where a crime is most likely to be perpetrated.

Another method of detecting and forecasting possible crimes is to listen in on phone calls made or received by a suspect. This is achieved by wiretapping a phone of a suspect.

Several law enforcement agencies use international mobile subscriber identity-catcher technology across the globe. A suspect's mobile phone can also be tracked through an international mobile subscriber identity catcher. This technology is also suitable for listening in on a suspect's phone calls. However, its use is controversial and has raised significant civil liberty and privacy concerns. As a result, international mobile subscriber identity-catcher technology is strictly regulated in many jurisdictions. This is also true for the use of drones in surveillance operations.

A significant disadvantage of conventional systems and methods of detecting and forecasting possible crimes is that these systems operate discretely.

Thus, there remains a need in the art for a system and method of tracking and forecasting possible crimes that will allow law enforcement agencies to combine all data gathered on a suspect and to focus its human and financial resources on a relevant suspect and location where a crime is most likely to be perpetrated. OBJECT OF THE INVENTION

It is accordingly an object of the present invention to provide a computer-implemented event tracking and forecasting system and method of using said system with which the applicant believes the above disadvantages would at least partially be addressed or which would provide a useful alternative to known systems and methods of tracking and forecasting an event.

SUMMARY OF THE INVENTION

According to a first aspect of the present invention, there is provided a computer- implemented event tracking and forecasting system comprising: a smart device having a display; an event tracking server having a memory module that stores data of a plurality of historical events, the data comprising at least type, time, and location identifiers of the historical events; and an event forecasting server that is configured to be in communication with the smart device and the event tracking server, the event forecasting server being operable to provide identifiers of a possible future event, the identifiers of the possible future event being based on a machine learning model that was obtained by processing the data of the historical events though a machine learning algorithm and the input data received from the smart device, the event forecasting module comprising: o a receiver module that is configured to receive input data from the smart device; o a memory module on which the machine learning model is stored, and which is further configured to store the input data from the smart device; o a processor module that is configured to retrieve the machine learning model and input data from the memory module and to process the input data through the machine learning model to extrapolate output data, the output data comprising at least type, time, location, and probability identifiers of possible future events that are based on the historical data; and o a transmitter module that is configured to transmit the output data of the processor module, in a format that is compatible with the smart device, to the smart device, wherein the output data of the processing module is displayed on the display of the smart device.

It will be appreciated that the term “smart device” means an electronic device that is connectable to other devices, networks, and servers and that can operate interactively with the said other devices, networks, and servers.

The smart device may be a smart device of a user of the computer-implemented event tracking and forecasting system. The user may be a law enforcement agent or an authorized official of a governmental agency. The smart device may be a personal computer, a laptop computer, a smart phone, a computer tablet, or a personal digital assistant. A plurality of smart devices and different types of smart devices may form part of the computer-implemented event tracking and forecasting system. There is provided for identifiers of an entity that is associated with a historical event to also be stored on the memory module of the event tracking server and/or the memory module of the event forecasting system. The entity may be a person, persons, association or associations.

The event tracking server may locate in a data center that is configured to be in communication with the smart device and the event forecasting server. The event tracking server may further be configured to be in communication with a plurality of smart devices and the event forecasting module. A plurality of event tracking servers may form part of the computer-implemented tracking and forecasting system. Each event tracking server may be operated and/or controlled by related or unrelated operators. The operator or operators may be a law enforcement agency or other governmental agency.

The plurality of historical events may also include identifiers and data of a person, persons, association, or associations associated with the historical events. The output data of the processor module of the event forecasting server may also include identifiers and data of a person, persons, association, or associations that are associated with the possible future event.

The time identifier stored on the memory module of the event tracking server may be the time at which a relevant event occurred. For example, the time identifier may be the calendar date and time of day on which an event occurred.

The type identifier stored on the memory module of the event tracking server may be an identifier of a type of historical event. For example, the type identifier may be an identifier of a committed type of crime. The type of crime may be any number of different crimes. Non-exhaustive examples of the type of crimes include drug crimes, theft, violent crimes, and illegal trafficking of prohibited goods.

The location identifier stored on the memory module of the event tracking server may be global positioning system coordinates of where a relevant historical event occurred. For example, the location identifier may be the global positioning system coordinates of a location where a crime was committed.

The smart device or smart devices, event tracking server or event tracking servers, and the event forecasting server may all be configured to be interactively connectable over a communication network. The communication network is typically the internet.

Data sent over the communication network and between the devices and servers may be encrypted for security purposes. In this regard, known encryption and decryption techniques may be used. For example, data may be stored in an encrypted format on the memory module of the event tracking server and the memory module of the event forecasting server.

Output data of the processor module of the event forecasting server may be sent in encrypted format from the transmitter module of the event forecasting server to the smart device, where it may, in turn, be decrypted and displayed on the display of the smart device. Alternatively, the output data of the processor module of the event forecasting server may be decrypted in the event forecasting server and sent in unencrypted format to the smart device, where it may be displayed on the smart device's display. The machine learning model may be trained by a method including the steps of:

(i) compiling a database or databases of historical events, the database or databases comprising at least type, time, and location identifiers of the historical events;

(ii) retrieving the data from the database or databases;

(iii) refining the data from the database or databases to provide refined data;

(iv) summarizing the refined data to provide summarized data; and

(v) processing the summarized data through a machine learning algorithm to provide a machine learning model that is suitable for processing input data that emanate from a smart device and extrapolating data therefrom that correspond to a possible future event.

The database or databases used in the training of the machine learning model may locate on the event tracking server or plurality of event tracking servers.

The machine learning algorithm may be a k-nearest neighbors algorithm, a k-means clustering algorithm or a linear regression algorithm.

According to a second aspect of the present invention, there is provided a method of using the computer-implemented event tracking and forecasting system according the first aspect of the present invention, the method including the steps of:

(i) establishing a connection between the smart device and the event forecasting server;

(ii) sending input data from the smart device to the event forecasting server; and (iii) receiving output data from the event forecasting sever, the output data having been obtained by processing the input data from the smart device through the machine learning model of the event forecasting server and being displayable on the display of the smart device.

The input data sent from the smart device to the event forecasting server may include data of an event obtained in-situ by a smart device user. This data may include identifiers of the type of event, details of persons or institutions associated with the event, and the calendar date and time of day that the event occurred. It will be appreciated that the afore is a non-exhaustive list of data that may be sent from the smart device to the event forecasting server.

Furthermore, the data sent from the smart device to the event forecasting server may include a request for output data comprising type, time of day, and location identifiers and a probability that a future event may occur in a suitable location. Still, further, the data sent from the smart device to the event forecasting server may include a request for output data on a person, persons, institution, or institutions associated with a possible future event.

BRIEF DESCRIPTION OF THE ACCOMPANYING DIAGRAM

The invention will now further be described, by way of example only, with reference to the accompanying diagram, wherein: Figure 1 is a schematic representation of an environment in which the computer- implemented event tracking and forecasting system according to the invention operates.

DETAILED DESCRIPTION OF THE INVENTION

According to the present invention, the computer-implemented event tracking and forecasting system is described below in more detail concerning a circumstance where the event being tracked and forecasted is a crime. However, it will be appreciated by persons skilled in the art that the computer-implemented event tracking and forecasting system of the present invention may also be used to track and provide predictions of events including earthquakes, acts of terrorism, and extreme weather conditions, and the like.

Regarding the accompanying diagram, in which numerals refer to like features, a computer-implemented event tracking and forecasting system according to the invention is generally designated by reference numeral 100 in Figure 1.

As shown in Figure 1 , the computer-implemented event tracking and forecasting system 100 comprise a smart device 200 having a display, an event tracking server 300 (hereinafter “crime tracking server”), and an event forecasting server 400 (hereinafter “crime forecasting server”).

The smart device 200, crime tracking server 300, and crime forecasting server 400 are connected interactively over a communication network 500. In Figure 1 , the communication network 500 is the internet. The smart device 200 is operable by a user 600. In the present case, the smart device 200 is a personal computer, and the user 600 is a law enforcement agent. However, the smart device 200 can also be a laptop computer, a smartphone, a computer tablet, or a personal digital assistant. A person that wishes to use the computer-implemented crime tracking and forecasting system 100 must first be authorized by the operator and/or controller of said system 100. In the present case, the operator and/or controller of system 100 is a law enforcement or governmental agency.

The crime tracking server 300 has a memory module 310 that contains data on a plurality of historical crimes. The plurality of historical crime data includes identifiers of the type of historical crimes, the calendar date and time of day on which the historical crimes were committed, and the global positioning system coordinates where the historical crimes were committed. The data on the plurality of historical crimes also includes data on an entity associated with the historical crimes. The entity may be a person, persons, association, or associations. Authorized persons, such as a law enforcement officer, may upload additional crime data to the memory module 310 of the crime tracking server 300 periodically or continuously.

The crime forecasting server 400 comprises: a receiver module 410 that is configured to receive input data from the smart device 200; a memory module 420 on which a machine learning model is stored and which is configured to store the input data received from the smart device 200; a processor module 430 that is configured to: o retrieve the machine learning model and input data received from the smart device 200 from the memory module 420; and o process the input data received from the smart device 200 through the machine learning model to extrapolate output data that is associated with a possible future crime, the output data comprising at least type of crime, calendar date and time of day, location, and probability identifiers of a possible future crime; a transmitter module 440 that is configured to transmit the output data of the processor module 430, in a format that is compatible with the smart device 200, to the smart device 200.

The output data of the transmitter module 440, which is received by the smart device 200, provides a forecast of the type, calendar date and time of day, location, and probability of a future crime occurring in a jurisdiction. The output data also includes identifiers and data on a person, persons, association, or associations that is/are associated with the possible future crime. The output data is sent to the smart device 200 in a format that is displayable on the display of the smart device 200.

The format of the output data is typically a report containing text. Ideally, the output data is displayed visually on a map on the smart device 200. Here, the output data is overlayed on the map. Visual elements that are displayed on the map typically include pins. The pins correspond to global positioning system coordinates of a location where possible future crimes may be committed, or historical crimes were committed. The type of crimes, the calendar dates and times of day, possible perpetrators of the said future crime, and the probability that a specific possible future crime may occur are also displayed on the map. Alternatively, the data mentioned above may be displayed on the smart device 200 after a user 600 of the smart device 200 has selectively selected a specific pin on the map.

The machine learning model that is stored on the memory module 420 of the event forecasting server 400, is trained by a method including the steps of:

(i) compiling a database or databases of historical crimes, the database or databases including type, time of day, location, and perpetrator identifiers of the historical crimes;

(ii) retrieving the data from the database or databases;

(iii) refining the data obtained from the database or databases to provide refined data;

(iv) summarizing the refined data to provide summarized data; and

(v) processing the summarized data through a machine-learning algorithm to provide a suitable machine learning model for processing input data emanating from a smart device and extrapolating data that correspond to a possible future event.

The database or databases are operated and controlled by a law enforcement agency.

Machine learning algorithms that are used to process the summarized data include a k-nearest neighbors algorithm, a k-means clustering algorithm, a linear regression algorithm and combinations of these algorithms. Once trained, the machine learning model is suitable for processing input data received from the smart device 200 and forecasting, utilizing pattern recognition and extrapolation process, output data, as described above, that is associated with a possible future crime.

The accuracy of the output data of the machine learning model increases when additional historical data is uploaded, over time, to the database or databases. The machine learning model is continually trained to provide updated machine learning models for forecasting a possible future event.

In use, the user 600 establishes a connection between the smart device 200 and the crime forecasting server 400 over the communication network 500. After a connection has been established, the user 600 sends input data to the event forecasting server 400. The input data typically takes the form of a request for a forecast of possible future crimes and/or a request for data on a suspect or suspects.

The user 600 may also connect the smart device 200 and the crime tracking server. After a connection has been established, the user 600 may also upload data obtained in-situ after an arrest has been made to the crime tracking server 300. Data uploaded to the event tracking server 300 is stored in the memory module 310 of the event racking server 300. In this manner, the database or databases whose data are used to train the machine learning model of the crime forecasting server 400 can be updated continuously. The crime forecasting server 400 processes the request received from the smart device 200 and determines whether the request is for data on a possible future crime and/or for data on a possible suspect or suspects. Data pertaining to the request are sent to the crime forecasting server 400 as input data. The input data is also stored on memory module 420.

The processor module 430 retrieves the machine learning model and input data from the memory module 420. It processes the input data through the machine learning model to provide output data as a forecast or result.

The output data pertaining to the request sent by the user 600 to the crime forecasting server 400.

For example, if the request is for data on a possible future crime, then the output data typically comprise identifiers of the type of possible future crime, a calendar date and time of day on which the possible future crime could occur, global positioning system coordinates of a location of where the possible future crime could occur, and a probability of the possible future crime occurring.

If the request is for data on a possible suspect, then the output data typically comprise identifiers of the suspect. Exemplary identifiers of the suspect include the name of the suspect, an identification number of the suspect, a passport name and number of the suspect, the nationality of the suspect, dates on which the suspect was arrested or apprehended, typical crimes that are associated with the subject and locations where the suspect was arrested or apprehended. Advantageously, the crime tracking and forecasting system 100 can combine a large amount of data from different databases and use said data to train the machine learning model used by and in the system 100. Any machine-readable data may be used to train the machine learning model. Data that may be used to train the machine learning model may also include video data, voice data, and image data. The trained machine learning model can then provide a forecast as described above. A law enforcement agent or agency can focus its resources on locations where a crime is most likely to be perpetrated by providing a forecast.