CH708144A2 - The process of selection of tourist recommendations rendered on a mobile device. - Google Patents

Abstract

The subject of the invention is a method for selecting tourist recommendations rendered on a user's portable device (1) in the form of audio and / or visual comments, comprising the following steps: the user selects an avatar (10) among several possible avatars (103); a personalized audio and / or visual commentary (12, 13) is returned to this user according to this selection.

Description

Technical area

The present invention relates to a method for selecting tourist recommendations returned to the user of a portable device.

State of the art

There are many tourist guides available in electronic form for example on smartphone, tablet or dedicated devices.

[0003] For example, electronic guides designed as a paper travel guide are known; the user can consult it by navigating to the page which concerns the points of interests likely to interest it. Such guides offer little interaction between the user and the software. A user who does not know anything about a particular region should consult a number of important electronic pages before finding the pages likely to interest them.

[0004] Electronic tour guides are also known, which offer tour itineraries, for example, in a city, museum or other tourist site. The route is sometimes displayed on a map. If the user's device is equipped with a satellite position receiver, such as a GPS receiver, it is possible to display the position of the user along the route. Some systems offer several courses of choice, including courses oriented around different themes such as "history", "shopping" or "nightlife" for example. The number of elective courses that can be offered is, however, necessarily limited. There is a need for systems and processes that allow for more courses, and more tailored courses for individual tastes. The selection of the most interesting route for a given user must however remain simple and it would not be acceptable for the users to have to make a selection in a long list of possible routes.

[0005] Electronic tour guides are also known that make it possible to provide audio or visual comments on users' devices. For example, audio guides are available in many museums to provide users with audio feedback as they visit; the speed of rendering these comments can be synchronized with the rhythm of the visit. Other known comment rendering systems are based on augmented reality solutions and allow users to view or listen to a comment associated with the objects filmed with their device. However, different users in one place do not necessarily want the same comments or recommendations. Some tourist sites, for example the center of tourist cities or amusement parks, can be associated with many different comments likely to interest different types of visitors, and among which a difficult choice must be made.

[0006] Tourist or commercial offer segmentation systems based on the use of forms that users must fill out are known. This operation is considered tedious by many users. Many users also do not like answering questions deemed inappropriate or indiscreet, for example questions about their age, sexual, religious or political preferences, their financial means, their origins etc. On the other hand, the storage of such data for many users is problematic in view of the data protection law of many countries.

Brief summary of the invention

An object of the present invention is therefore to provide a method and a system for automatic selection of tourist comments returned to a user on a portable device.

According to the invention, these objects are achieved in particular by means of a method for selecting tourist recommendations rendered on a portable device of a user in the form of audio and / or visual comments, comprising the following steps: the user selects an avatar from several possible avatars; a personalized audio and / or visual commentary is returned to this user based on this selection.

The avatar is preferably a graphic representation, more or less realistic or in a more cartoon style, that the user selects and adapts to represent themselves.

An advantage of this method is that the definition of preferences for tourist comments of the user is fun and graphic.

Another advantage is that this selection is non-explicit. The user is not forced to choose from a long list of courses or comments available, but simply choose and possibly customize an avatar.

Another advantage is that this selection is considered less intrusive. A user will probably be less reluctant to indicate by his avatar an age range than to provide similar information in a form. A user will have even less reluctance to represent themselves by a different person by means of an avatar who does not represent him, than to cheat in a form.

[0013] The correspondence between this avatar and the comments offered is not explicit. However, tests show that the appearance that a user wants to give to his avatar says a lot about his personality, his hobbies and his preferences. This is particularly the case if the avatar is supposed to represent the user, or the character that the user would like to play.

For example, the size of an avatar can determine approximately whether it is a child or an adult. His sex can be determined from his appearance. Clothes probably reveal if he is on a business or private trip, and also say a lot about his personality and style. The accessories he can equip, such as a hip hop cap or a violin, often determine his main points of interest.

In one aspect, the method relates to an automatic and / or manual categorization of users in different segments, each segment being associated on the one hand with a set of comments or recommendations, and on the other hand with one or more characteristics of avatars chosen by a majority of users in this segment to represent themselves.

[0016] The selected comments may include comments on points of interest near the user.

The selected comments may include recommendations of points of interest to visit, or route recommendations.

The avatar can be selected by the user from a list of users available. For example, a first user interested in culture may choose an avatar representing an artist; a tourist interested in sport will choose a sports avatar; etc.

The user can choose different avatars at different times. A user can thus express several different personalities at different times.

The user can customize the selected avatar, for example by means of accessories or by modifying its characteristics. Future audio and / or visual comments are selected based on this customization. The number of possible avatars can be important depending on the customizations. For example, a person interested in rock music can provide his avatar, whatever it is, an electric guitar. Future comments, including recommendations for places to visit, will henceforth take this taste expressed for rock music.

The avatar selected and optionally personalized by the user can also be automatically customized according to the course previously traveled by said user. For example, a user who has entered one or more times in a rock room may have his or her avatar automatically decked out with a guitar. Automatic customization can be performed without any action from the user, or require user validation. The future audio and / or visual comments selected for this user are based on this automatic customization.

Automatic personalization may depend on manual customizations made by other users who have traveled a similar course. For example, the system can detect that users entering a particular room share a common feature of their avatar, such as a high proportion of these users have an avatar with a guitar, to decide to automatically propose this guitar or a personalization trait common to future users present at the same place or who travel a similar course.

The automatic personalization can be temporary and depend on the tourist or commercial offer available nearby. For example, a user near a restaurant selling hamburgers may observe that his avatar is momentarily wearing a cap with the logo of that restaurant.

In this way, the user can be informed that it is now likely to review the audio or visual commentary subsidized, for example advertising. He may choose to decline this targeted advertising by refusing or removing the automatic personalization feature that has been proposed to him.

[0025] A catalog of audio and / or visual comments can be stored in a database. The database may be local in the user's device. The comment database can be stored remotely in a server accessible by the user's device. A database can be in the user's device and synchronized with a remote database in a server. The server can be one or more machines, including a virtual server or a cloud.

At least some comments may be associated in the database at geographical locations. For example, an audio and / or visual commentary on the Geneva water jet may be associated with a round or rectangular zone around the Geneva water jet, or with all the geographical areas from which the jet of water water can be seen or from which it is easy to get there.

At least some comments may be associated in the database with characteristics of an avatar. For example, a commentary in the form of a suggestion to visit a particular rock room may be associated with the "electric guitar" feature that adorns some avatars.

An avatar characteristic may also be associated with a vocabulary, expressions, a voice or a particular accent. An identical (or different) comment can be rendered differently, for example with a vocabulary, expressions or a different voice, depending on the chosen avatar characteristics.

Comments returned to the user depend both on its geographical location and the characteristics of its avatar. For example, two users in the same location who have selected or for whom the system has automatically selected different avatars will have different comments. Two users with the same avatar or similar avatars will have different comments if they are not in the same place. The selected comments are both geolocated and depending on the preferences of the user visualized and selected by means of his avatar.

Brief description of the figures

Examples of implementation of the invention are indicated in the description illustrated by the appended figures in which:

FIG. 1 illustrates an example of an avatar displayed on a portable user device.

FIG. 2 illustrates an example of a tourist recommendation displayed on a portable device in the form of a route recommendation.

FIG. 3 illustrates an example of a tourist recommendation displayed on a portable device in the form of audio commentary returned to the user when he is in front of a monument.

FIG. 4 schematically illustrates a system according to the invention.

FIG. 5 is a flow diagram of an exemplary method according to the invention.

FIG. Figure 6 illustrates a module for mapping avatar characteristics to tourist comments stored in a database.

Example (s) of embodiment of the invention

A system for implementing the invention is schematically illustrated in FIG. 4. It comprises a plurality of portable devices of users 1, for example smartphones, tablets, PDAs, interactive glasses or dedicated devices. Several different type devices can be combined in the system. The implementation of the system involves the installation in portable devices 1 of dedicated software, for example applications, or an internet browser to access web pages.

The portable devices 1 are connected via a wireless network 2, for example a cellular network, a wireless network, and / or the Internet, to a server 3 comprising a processing system 30 and a database 31.

The database 31 stores a catalog of audio and / or visual comments 311, as well as avatar characteristics 310. In a variant, the database 31 can also be local and stored in the portable device 1 of the user. The database 31 can be in the portable device of the user 1 and synchronized with a remote database in the server 3. The server 3 can be constituted by one or more machines, including a virtual server or a server. cloud.

The method of the invention will now be described with reference to FIG. 5 and with the help of other figures if necessary.

During step 100, the user starts a software or a dedicated application on its portable device 1, or connects with its portable device 1 to a dedicated web page providing its identification.

In step 101, the software or application in the portable device of the user, or software running in the server 3, checks if an avatar is available to the identified user.

If an avatar is already available, the software or the application optionally proposes the user to modify or customize it during the optional step 102. The customization may include for example modifications of the avatar such as changes in size, physical appearance, clothing, accessories, etc. It is also possible to offer the user a choice of several avatars he would have selected during previous sessions, or the creation of a new avatar in addition to the avatars previously saved.

If no avatar is available, the software or the application proposes to the user during step 103 to create a new avatar, The creation of a new avatar can include a selection from a list of avatars available. The available avatars correspond for example to different clearly defined personalities, for example "girl about 6 years old", "young man amateur rock music", "adult interested in classical culture", etc. It is also possible to propose avatars corresponding to personalities of the world of the spectacle or politics, for example to visit a city like Georges Clooney, etc. It is also possible to propose avatars corresponding to heroes of fairy tales, fantastic stories, characters of comics or manga, according to the activities and the places. It is also possible to allow a user to use his photograph as an avatar or part of an avatar.

According to an important feature of the invention, the avatar can also be used to indicate disabilities of the user. A visually impaired person may, for example, equip his avatar with glasses and / or a cane. A person with reduced mobility will be able to equip his avatar with a wheelchair. The itineraries subsequently offered to this user take into account the handicaps thus reported. Standardized coded accessories can be used to indicate different disabilities.

The avatar can also be used to indicate the user's food preferences, for example a vegetarian or kosher diet, or a food allergy. The avatar can also be used to indicate the ethnic type of the user. The avatar can also be used to indicate the profession of the user, for example by equipping the character with accessories specific to a profession. The avatar can also be used to indicate a user's sexual orientation. Other categories of users sharing common tastes can thus be grouped around accessories or common features of their avatars.

In a variant, the creation of a new avatar during step 103 may be based on a series of questions or successive choices. For example, a user may have to choose successively the size of his character, the color of his hair, his clothes, etc. using a graphical choice interface or textual questions.

In a variant, the creation of a new avatar during step 103 may be based on the possession of a real physical object associated with an avatar. For example, the user may purchase a stuffed toy, bag, book or other object representing the avatar, and enter into the software or application a code, for example a digital code, a barcode, a datagram, etc. on this object to bring up a corresponding virtual avatar on the screen.

In step 104, the user can then customize the created avatar during the previous step 103, in the same way and with the same freedom as in step 102.

The avatar thus created and / or personalized corresponds to an alter ego of the user, supposed to represent in a visual form his personality or in any case the personality of the traveler he wants to be. An example of an avatar 10 with an accessory 100 (here a guitar) is shown in FIG. 1. Arrow 11 allows you to choose another avatar. The characteristics of this avatar are advantageously stored in memory, in the portable device 1 and / or in the remote server 3. This avatar can evolve, as will be seen later, with or without user interaction, so as to to define more and more precisely the personality of the user.

The following steps illustrate examples of steps that can be implemented to use this avatar to automatically select tourist comments returned in auditory and / or visual form to the user.

During the optional step 105, the position of the portable device of the user is determined. This location can be performed for example by means of a satellite GPS location receiver GPS, Magellan or equivalent, and / or using a network of terrestrial transmitters, for example cells of cellular networks or points WIFI access, and / or image recognition algorithms to locate the device using images captured by an integrated camera. Different geolocation methods can be combined, simultaneously or according to their availability; for example, it is possible to use an inaccurate localization by cells of cellular networks at any time, and to refine the geolocation when a GPS signal is available or that a recognizable scene is captured with the camera on board. A series of successive positions can be determined and recorded in order to determine in the portable device 1 and / or in the server 3 the trajectory of the user.

During the optional step 106, the last determined position of the user, or recent trajectory, is used to refine the knowledge of this user, and automatically adapt his avatar based on this new information. For example, a user who goes to a rock club one or more times will be able to see his avatar automatically with a guitar or other attributes related to the place visited. The avatar of a user on the beach may be dressed in beach clothes, at least temporarily or even when he left the beach.

This adaptation of the avatar can be performed automatically by the server 3, or by a software in the portable device 1. A validation of the user can be requested, or the modification can be done directly without the need to validate. Several adaptation options can be proposed from which the user can choose.

The modifications of the avatar proposed to the user may depend, for example avatars chosen by previous visitors of the same locations. For example, software running in the server 30 can determine the original avatar characteristics shared or accepted by a large number of visitors to a site, and offer these original features to future visitors. The software can for example determine the type of clothing style that is more common among the avatars of opera visitors than among the general population, and suggest or directly apply this style for the avatars of future visitors. A suggested change can be made at the first visit of a location, or after several visits to the same location or similar locations.

The modifications of the avatar proposed to the user may be sponsored by institutions or companies. For example, an amusement park can pay the operator of the server 3 a fee so that visitors to his park are equipped during the visit, or out, an accessory reminding them to visit the park.

The changes of the avatar proposed to the user may be semi-permanent, or temporary. Modifications are said to be semi-permanent if they are maintained as long as they have not been deleted by the user or replaced by other changes that make them obsolete. Temporary changes for example have a limited life before disappearing, for example a predefined time duration, or appear only as long as the user is close to the location associated with this modification.

The proximity between a user and a location can be determined for example on the basis of a geographical distance. For example, a user may be considered to be near a given location if he is within x distance of that location. The distance associated with different locations may be variable. It is also possible to consider that the user is near a location if he is in the same cell of a cellular network of this location, or in a WIFI network associated with this location.

The proximity between a user and a location can be determined for example by taking into account the interest that this user has shown for this location. For example, it is possible to consider that the user is near a location if he has seen or listened to a comment associated with that location, or if he has otherwise manifested an interest in that location. The efforts made by a user to go to a given location can also be automatically taken into account by the software in the server 3. For example, a user who chooses among several possible recommendations to visit a rock club at the other end The city, while other recommendations more frequently followed by other users were less far-reaching, is likely to demonstrate above-average interest in rock music.

The proximity between a user and a location can be determined for example based on the time spent by a user at this location. Thus, a user who only goes past the opera may be classified as not very interested in the opera, while a user who enters the opera and spends several hours there reflects a proximity and interest higher. Minimum durations may be associated with different locations to determine sufficient proximity.

In the course of step 107, an audio and / or video commentary is returned to the user on his portable device 1. The selection of the restored comment is performed by software executed in the portable device 1 and / or server 3, taking into account both the current position of the user and the characteristics of his avatar. It is also possible to take into account the past trajectory of this user, that is to say at least one previous location, in order to avoid giving the user a comment about a location or a point of interest. interest he has already visited or for which he has already heard or seen the comment.

It is possible to simultaneously return to the user different comments or offers. These different comments can be sorted according to their supposed relevance for the user according to the characteristics of his avatar. The choice of the selected comment can be used to adapt on the one hand the avatar of this user, on the other hand to propose modifications of avatar to the future users making the same choice.

Local filtering of the recommendations can be performed to filter recommendation suggestions based on data available in the user device. For example, suggestions for routes or visits may take into account data in the user's calendar, or other data available in the user's device 1.

The comment returned to the user may for example correspond to a route suggestion, as shown in FIG. 2. The route suggestion may be displayed as a map 12 as shown in the figure, or audibly rendered. The map can indicate the current position of the user and / or a view of the environment from the position of the user.

The comment rendered to the user can for example correspond to audio comment, as shown in FIG. 3. This audio comment can be triggered automatically or after a confirmation from the user when it is near a point of interest associated with such comment. This comment may for example be rendered when an image 13 of the commented point of interest is captured by the camera of the portable device 1, recognized and displayed on the screen of this device. Known algorithms in the field of augmented reality can be implemented to align the images captured by the camera of the device 1 with reference images in the database 31, and to trigger the return of a comment when point of interest corresponding to the interests of the user as they can be deduced from his avatar is thus recognized on the image.

The natural or synthetic voice used for a restored audio comment may depend on the selected avatar. For example, a user who selects an avatar for a fairy-tale character will be given, for example, comments in the form of tales telling him about the city or points of interest of the city visited. A user who chooses a known personality will have comments rendered with the voice of that personality.

During the optional step 108, the user can give an evaluation of the comment that has been returned to him, to indicate whether he has appreciated and / or if this comment corresponds to his tastes. For example, he can accept or reject a route proposal, or give a note on an audio or visual comment rendered in front of a point of interest. This evaluation can be used to modify the accessories associated with this comment for future visitors. For example, a user represented by an avatar with a guitar and who agrees to visit a given location, possibly giving it a positive rating, will be detected by software that will likely offer such a guitar to the avatars of future users visiting this site.

This evaluation can also be used to generate an automatic modification, or suggestion of modification, of the avatar of the user. For example, a user who systematically refuses to visit places devoted to rock, or who visits them briefly and leaving negative evaluations, may be able to remove his guitar, or offer to remove it.

The process then returns to step 105.

The database 31 preferably stores a list of avatar characteristics 310, as illustrated in FIG. 6, as well as a comment list 311. Each comment can be associated with a given location or point of interest, so that it can be returned to users near that point of interest. The concept of proximity is described above, and can be determined in different ways.

A software module 300 executed in the user device 1 and / or in the server 3 then establishes a correspondence between the avatar characteristics 310 (or combinations of avatars characteristics) and the comments (including the recommendations) 311. This correspondence can be binary, or preferably take several values to indicate a more or less strong correspondence. For example, a bike as an avatar feature will have a high value match with bike shops or bicycle stadiums, and a lower value match with general sport stores.

The software module 300 may be based on a linear classifier, for example on the technology of support vector machines (SVM), in order to classify each avatar characteristic in one or more categories each corresponding to a given comment or to a group of comments. An SVM model of artificial learning can be associated with each comment or feature of an avatar. A new SVM template is created automatically by the software module if a new avatar feature is created or if a new comment is introduced. Each SVM model learns to offer offers based on the data captured by each avatar. The software module can also be based on a neural network.

Thus, each time a user requests an avatar, whether during an activity, by going to a particular place, or by making a particular choice, the avatar evolves and enriches, while the software 300 learns new associations, or refines existing associations, between avatar characteristics and user requests or preferences.

The system is preferably able to learn for itself these matches, depending on the choices of users who have avatars with such and such characteristics. An artificial learning algorithm can be implemented for this purpose.

New characteristics or combinations of characteristics can be created at any time, and associated little by little with different locations. Users can draw new features of their avatar with a suitable computer drawing tool. An operator to also forcefully create matches between avatar characteristics and comments related to different locations. This is particularly useful when starting the system to provide immediate personalized feedback to users who have chosen different avatars.

The method thus realizes a segmentation of tourism offers automatically based on the graphical characteristics that make up the avatars of different users, their geographical location, and possibly other elements. The information and the choices made by the avatars are grouped by identical avatars, or characteristics of identical avatars, or near graphs.

The method also allows each user to quickly endorse another personality, the time of a visit, without having to fill a tedious form, and simply changing avatar or manually changing his avatar.

Several users can be represented, at least temporarily, by the same avatar or by avatars sharing a common feature. This makes it possible, for example, to offer common or simultaneous choices to several users, for example during an event. It is thus possible to create gathering places around a physical avatar, or an avatar projection around which the persons concerned can gather. For example, in a festival or sporting event, one can imagine projecting the common avatar onto a monument or part of a landscape where users are invited to gather. It is also possible to distribute a crowd in different places, for example during a trade show, etc.

Claims (10)

A method for selecting tourist recommendations rendered on a portable device (1) of a user in the form of audio and / or visual comments, comprising the following steps: the user selects an avatar (10) from among several possible avatars (103); a personalized audio and / or visual commentary (12, 13) is returned to this user according to this selection. The method according to claim 1, comprising a step (102, 104) in which said user personalizes said avatar (10) by means of accessories or by modifying its characteristics, the future audio and / or visual comments being selected. according to this customization. 3. The method according to claim 2, comprising a step (106) during which said avatar is automatically personalized according to the course previously traveled by said user, the future audio and / or visual comments being selected according to this personalization. 4. The method of claim 3, wherein the automatic customization depends on manual customizations made by other users who have traveled a similar path. 5. The method according to one of claims 3 or 4, wherein the automatic customization depends on the tourist or commercial offer available nearby and therefore varies when the user moves. The method according to one of claims 1 to 5, wherein a catalog of audio and / or visual comments (311) is stored in a database (31), at least some comments being associated in the database (31) with geographical locations, at least some comments being associated in the database (31) with avatar characteristics (310), the comments returned to said user depending both on his geographical location and the characteristics of his avatar. The method of claim 6, comprising an automatic association step between comments (311) and avatar features (310). 8. The method of claim 7, said automatic association implementing vector support machines. 9. The method according to one of claims 7 or 8, said association depending on the avatar accessories of previous users having visited a given location. A tangible computer medium containing computer code to be executed by a computer processing system (1, 3) for implementing the method of one of claims 1 to 9 when the computer processing system executes said computer code. .