Computer implemented method, electronic device and digital platform for determining a user-associated desired prop-erty of a physical item

Abstract

The present disclosure relates to a computer-implemented method, to a device carrying out the method and to a digital platform, which implements the method, the method being configured for determining a user associated desired property (104) of a physical item (102), comprising to select (S1) a virtual item (103x), for which the user associated desired property (104) is to be determined and to determine (S3) the user associated desired property (104) of the physical item (102) represented by the selected virtual item (103x), wherein the user associated desired property (104) is determined by a user associated property algorithm (150), which uses at least one relevant item property (105) from a virtual item group (111) of at least one other user (100) from a item group database (300).

Description

[0001] COMPUTER IMPLEMENTED METHOD, ELECTRONIC DEVICE AND DIGITAL PLATFORM FOR DETERMINING A USER-ASSOCIATED DESIRED PROPERTY OF A PHYSICAL ITEM

[0002] FIELD OF THE DISCLOSURE

[0003] The present disclosure relates to a computer-implemented method for determining a user associated desired property of a physical item, to a data processing device comprising a processor configured for carrying out the computer-implemented method for determining a user associated desired property of a physical item, to a digital platform implemented as a computer program comprising instructions, which, when the program is executed by a processor, cause the processor to carry out the computer implemented method for determining the user associated desired property of the physical item and to a computer-readable medium comprising instructions, which, when executed by a computer, cause the computer to carry out the computer-implemented method for determining the user associated desired property of the physical item.

[0004] BACKGROUND OF THE DISCLOSURE

[0005] A high number of items, like cloth items, books etc. are purchased online and are sent via a delivery service to the customer. Many of the customers order for one desired item a plurality of available sizes in order to try them at home and to choose at home the desired size best fitting to the individual user. The other items or all items, in case the right size was not included, are sent back to the seller. This creates many delivery costs because all the items need to be sent to and back from the customer. Further, this shipment or delivery creates a lot of CO2 due to the shipment itself and due to the fact that some items, which are sent back may no longer be in a condition in which they can be resold. In addition, the currently available system creates a lot of administrative work for the online retailer because he needs to commission and deliver the large order per customer, he needs to overlock the return shipment, he needs to process the items which have been sent back and he needs to put them back in stock if they can be resold otherwise he needs to dispose them. Overall the entire process is not as efficient as possible and produces a lot of unnecessary CO2 emission.

[0006] In case the returns of all items ordered online in Europe could be reduces by e.g. 30% the CO2 savings would be huge. Further, a system, which enables the possibility to sell and buy second hand articles, and which avoids that items are sent back (and can due to bad condition no longer be sold) could further reduce the CO2 emission drastically. E.g. the CO2 savings from purchasing second-hand items are e.g. for a dress: 2.72kg, a jacket: 6.42kg, a T-Shirt: 1 .47kg, a pullover: 4.62kg, a shirt / blouse: 2.05kg, a pant: 4.07kg, a skirt: 1 .83kg, shoes: 6.7kg.

[0007] The target should be to reduce the number of items, which are sent back by the customer. Avoiding returns saves an average of 0.85 kg of CO2 per shipment. This would reduce the CO2 emissions during the shipment process, the CO2 emissions required for returning and the CO2 emissions for production of returned and unsellable items. The main reason for sending received items back by customers is that the items do not fit the consumer’s body size.

[0008] If a person buys e.g. shoes in a store I can try as many shoes as possible from different brands having different sizes until finally the shoes, which perfectly fits is found, the size itself is not that relevant. This is online not possible. Currently the customer orders for a desired type of shoe a plurality of sized for trying them at home. The shoes, which do not fit, are sent back. Even if the customer knows its normal shoe size, different brands build the shoes slightly different such that it is normal that for one brand or shoe type the customer needs one size and for another brand or shoe type the customer needs another size. Overall, it is required that a plurality of shoes are delivered in order to find the perfect match, which creates the unwanted high CO2 emissions. This should be reduces in particular by reducing the returns.

[0009] SUMMARY OF THE DISCLOSURE

[0010] It is an object of the present disclosure to provide a computer-implemented method, a data processing device, a digital platform implemented as a computer program and a computer readable medium, which address the above-mentioned problems or issues. Further, it is an object of the present disclosure to provide a computer-implemented method, a data processing device, a digital platform implemented as a computer program and a computer readable medium, which are configured to determine a user associated desired property of a physical item, in particular a user associated size of a clothing item, thereby in particular reducing the returns of items and wrong allocation of items to users / people or a group of users / people.

[0011] According to the present disclosure, these objects are addressed by the features of the independent claims. In addition, advantageous embodiments follow from the dependent claims, figures and the description.

[0012] According to the present disclosure a computer-implemented method for determining a user associated desired property of a physical item is specified. The user associated desired property is for example a size of a clothing item, in another embodiment the user associated desired property may be a disease, a mental disorder, a specific behavior or any other property, which is assignable to one or a plurality of users. Further, the users are not limited to human beings. In the context of the present disclosure, the user may comprise animals, plants, artificial intelligences, products etc. The computer-implemented method may comprise the following steps of (the order may vary):

[0013] Selecting, by an access user, a virtual item, representing the physical item from a selectable set of virtual items stored in an item database, for which the user associated desired property is to be determined. The access user is the user for which the user associated desired property should be determined. The virtual item is a virtual copy or digital twin of the physical item, which the access user may want to buy. The physical item is e.g. a shoe of brand A. The virtual item is a virtual copy of that shoe of brand A stored in the item database. The item database collects and provides the possibility for the access user to select the respective desired virtual item. In other words, the item database is a collection of virtual items (products) or digital twins of physical items. The access user may select via an application on his smartphone or via a computer from the item database the desired virtual item. Further, it is also conceivable that third party applications pro- vide via e.g. an online store the selectable set of virtual items. E.g. the request comprising information on the desired virtual item is received by the computer-implemented method. In this case, the virtual items may be selected indirectly by the access user.

[0014] In a further step, an item group database is provided, or access to an item group database is provided. The item group database stores for the access user and a plurality of other users a respective virtual item group, representing a physical item group, the virtual item groups comprise at least one virtual item of the same item category as the selected virtual item, each virtual item being stored with a plurality of respective relevant item properties. The item group database may correspond at least partially with the item database. In other words, the item group database and the item database may be one large database, which provides all the required functionalities. The databases may be hosted on a local or cloud based server providing access for the computer implemented method such that the required information or data may be updated and / or retrieved from the databases. The item group database stores for the access user and the plurality of other users the virtual item group, representing the physical item group. The physical item group is e.g. a shoe wardrobe or a wardrobe of shirts or trousers of the respective user. The virtual item group is the digital twin of the respective physical wardrobe. E.g. the virtual item group of the shoe or sneaker category, of one user comprises fife pairs of shoes of e.g. fife brands and with the respective shoe size. The fife shoes are in this example the virtual items and the respective brands and the respective shoe sizes are the relevant item properties. At least some of the stored virtual items of virtual item groups of the user may correspond to the virtual items selectable from the item database. E.g. after the step of selecting the desired virtual item for which the user associated desired property should be determined, the method, in particular the application or a processor, accesses the item group database and retrieves information from the item group database for determining the user associated desired property e.g. the best fitting shoe size.

[0015] In a further step, the user associated desired property of the physical item represented by the selected virtual item is determined, in particular by the computer implemented method running on the processor. The user associated desired property is determined by a user associated property algorithm, which uses at least one relevant item property from the virtual item group of at least one other user from the item group database. The user associated property algorithm is an algorithm, which accesses the respective databases or which uses information from the respective databases to determine the user associated desired property. The user associated property algorithm uses the relevant item property, e.g. the brand or the size of the virtual items of the virtual item groups of at least one other user stored in the item group database.

[0016] By using information of other users, other people, it is advantageously possible to determine the user associated desired property of the respective physical item. E.g. physical items, which are used in a matching manner, by one user may perfectly match to the access user if they further share other virtual items in their respective virtual item group. By using information from other users, it therefore possible to determine the desired user associated property of a physical item, e.g. the matching size of a pair of shoes with a high accuracy. It is therefore only required to buy only this particular shoe. This reduces the commissioning requirements for the online retailer, the shipment costs and the return costs, because only one pair of shoes is shipped to the access user. Overall, the present disclosure enables that the CO2 emissions are drastically reduced.

[0017] In an embodiment, the user associated property algorithm, running e.g. on the processor of the device the access user uses or on a server, further uses relevant item property data from the virtual item group of the access user for determining the user associated desired property. The user associated property algorithm may use e.g. the relevant item properties, like brand and size of e.g. the shoes of the access user, and I or additional information, like standard size, body dimensions or other body weight etc. from the assess user, stored e.g. on a user profile, in particular on the item group database. This may further increase the accuracy of the determination of the user associated desired property.

[0018] In an embodiment, the user associated user associated property algorithm comprises or is implemented to comprise at least one of the following steps.

[0019] Searching in directly matching virtual item groups of the item group database for the selected virtual item. The searched directly matching virtual item groups of the other users comprise at least one other virtual item, which is also comprised in the virtual item group of the access user having the same relevant item properties. The virtual item group, as virtual shoe wardrobe of the access user, may comprise two pairs of shoes (shoe 1 of brand A and size 10, shoe 2 of brand B and size 9.5). The directly matching virtual item groups of other users also comprise either shoe 1 or shoe 2 if possible even both, which would provide an even higher accuracy. The virtual item groups of the other users further should comprise the selected virtual item (e.g. shoe C) as selected by the access user.

[0020] In case the selected virtual item is found in at least one directly matching virtual item group the user associated desired property of the selected virtual item for the access user is determined, by extracting the respective desired relevant item property from the found selected virtual item from the at least one directly matching virtual item group of the respective other user. E.g. the virtual item group, e.g. the virtual shoe wardrobe of another user, which comprises the one directly matching virtual item with the virtual item group of the access user and which further comprise the selected virtual item (shoe C in size 9.5) is accessed by the user associated property algorithm and the user associated property algorithm retrieves the size of the selected virtual item from the other user (e.g. size 9.5) as user associated desired property and may show the respective size to the access user. In other words, if one specific pair of shoes e.g. of brand A size 10 is present in the virtual item group of the access user and the other user and the other user also has the selected shoe e.g. shoe C (virtual item) in his virtual shoe wardrobe with the size 9.5, it is highly likely that the respective shoe C would also fit the access user in size 9.5. The size 9.5 is thereby determined as the user associated desired property and e.g. presented to the access user as result. The access user may than buy only the shoe C in size 9.5, as physical item, which should fit with a high likelihood well. The order size and the return shipment is thereby reduced drastically.

[0021] In a further embodiment, the user associated property algorithm is configured to determine the user associated desired property of the selected virtual item in case the selected virtual item is found in a plurality of virtual item groups, preferably directly matching virtual item groups, by extracting and using the desired relevant item property of the found selected virtual item from the plurality of virtual item groups of other users. In this case, a plurality of virtual item groups, e.g. a plurality of virtual shoe wardrobes, comprises the selected virtual item. The accuracy of the determination is increases if the relevant item properties of the plurality or at least a portion of the plurality is selected and used by the user associated property algorithm to determine the desired user associated property. Further, in case a plurality of direct matching virtual item groups comprise the selected virtual item the relevant item property (e.g. the size) is extracted from the different virtual item groups and in combination used (e.g. via averaging or weighted averaging) for determining the user associated desired property. This further increases the accuracy of the determination.

[0022] In a further embodiment, the computer-implemented method may be configured to provide the possibility of adding, by a user, a new virtual item to the item database, thereby updating the item database, the newly added virtual item becomes thereby selectable by at least one user. The user or a retailer may for example add via an application or a computer a new virtual item to the item database. E.g. new clothes are constantly designed, which need to be added to or removed from the item database. This may be done by the retailer, the company producing the items, by a user or by anyone who has access and the respective permissions to and for the item database.

[0023] The computer-implemented method may further be configured to provide the possibility for adding a new virtual item group of a new or existing user, in particular for another item category, and I or for updating an existing virtual item group by adding a new virtual item or by deleting at least one virtual item, thereby updating the item group database. One user may have in his profile a plurality of virtual item groups in particular for different item categories. E.g. one virtual item group to virtually represent his sneakers, one for his boots, one for his t-shirts one for his business shirts etc. It is possible to add for one user a new virtual item group for creating the respective virtual twin of the physical item groups. Further, it is of course possible for the respective user to amend a specific virtual item group by adding or deleting an item e.g. the respective item does not fit any longer. This may trigger a new classification of the respective virtual item group.

[0024] In an embodiment, the computer-implemented method is, in particular in case the selected virtual item is not found in at least one directly matching virtual item group, configured to provide, to provide access to the item group database, in which each virtual item group is classified in an item group cluster based on the virtual items of the respective virtual item group and the respective relevant item properties of the virtual items. In this embodiment, the computer implemented method running on the processor, accesses or has access to the item group database, in which each virtual item group is classified in an item group cluster. In other words, the item group database comprising the virtual item groups further comprises a cluster classification for the virtual item groups. The cluster classification is determined in that each item group cluster comprises only virtual item groups, which are free of a conflict with respect to each other, wherein a conflict is determined in that two virtual items of different virtual item groups (of the same item group cluster), which have one relevant item property in common have a varying other item property. If this is not the case, the respective item group cluster is free of a conflict. One item group cluster comprises e.g. three virtual item groups each comprising virtual items of e.g. shoes. The different shoes are from different brands and may have different sizes as relevant item properties, but the cluster does not comprise two shoes of the same brand having different sizes. This makes the cluster conflict free.

[0025] According to this embodiment, the associated property algorithm may be configured to search in the item group cluster, in which the virtual item group of the access user is classified, for the selected virtual item. This embodiment is in particular used in case the selected virtual item is not found in a directly matching virtual item group. Each virtual item group is according to this embodiment classified in clusters and the selected virtual item is searched in this cluster of the access user.

[0026] The user associated property algorithm is, in case the selected virtual item is found in the item group cluster, in which the virtual item group of the access user is classified, configured to determine the user associated desired property of the selected virtual item, by extracting the respective desired relevant item property from the found selected virtual item in the item group cluster. In other words, in case the selected virtual item is found in a virtual item group of another user classified in the same item group cluster the desired relevant property is extracted from that found virtual item and provided / displayed to the access user as the user associated desired property. In case the selected virtual item is found in a plurality of virtual item groups of the same cluster, the accuracy of determination will be higher. Clustering and searching in the respective cluster further increases the accuracy of determining the user associated desired property. In a further embodiment, in case the selected virtual item is found in the item group cluster, in which the virtual item group of the access user is classified, the user associated property algorithm is further configured to determine the at least one other user in whose virtual item group the selected virtual item is found and use this information for the determination of the user associated desired property e.g. by weighting the activity, actuality, and I or number of items of this user.

[0027] Further, the user associated property algorithm may be configured for searching if at least one virtual item group of a different item category of the at least one other user is also classified in the same item group cluster of this item category of the access user and retrieving relevant item properties of matching virtual items of this item group cluster. This may further increase the accuracy of determination. Further, the user associated property algorithm may be configured for determining the user associated desired property of the selected item by further using the retrieved relevant virtual item properties of the matching virtual items of this item group cluster. According to this embodiment, information of different virtual item groups are additionally used for determining the user associated desired property. E.g. in case the selected virtual item is a shirt, and the selected virtual shirt is found in a virtual item group of another user classified in the same cluster, further information of the other user may be used for determining the user associated desired property. It is looked if other virtual item groups of this user (e.g. the virtual item group representing the trouser wardrobe) has a directly match or is classified in the same item group cluster (trouser cluster) as the virtual item group of this item category (trouser) of the access user. This information is further used to determine the desired user associated property, e.g. the size and fit of the selected shirt of the access user. This further increases the accuracy of the determination of the user associated desired property.

[0028] In a further embodiment, the computer implemented method comprises the step of updating the classification of the virtual item group in the available item group clusters, in case the respective virtual item group is amended by adding a new virtual item or removing a virtual item from the virtual item group. A respective updating algorithm may be used, which runs on a processor via an application and I or the database, or the respective devices or application being capable of updating the database. The item group clusters need to be updated, preferably automatically, in case the comprising virtual item groups are amended, in particular by removing or adding virtual items. E.g. the best fitting clothing items may vary over time e.g. due to changing body dimensions (e.g. during a pregnancy). It is therefore necessary to update the classification of the virtual items if e.g. a new virtual item (a shirt of a larger size) is added to the virtual item group. Adding a shirt of a larger size to the respective virtual item group may trigger a conflict, because the same shirt may be already part of another virtual item group of another user of this cluster, because now there are two items of the same brand but of different sizes in this cluster. By updating the classification of the virtual item groups in the cluster, it is possible to keep the respective clusters conflict free.

[0029] In an embodiment, updating the classification may comprise, in case the amended virtual item group is not jet classified in any of the available item group clusters, to add the virtual item group to an already existing item group cluster if the amended virtual item group fits in one of the already existing item group clusters without creating any conflict, or the amended virtual item group is added to a newly created item group cluster if the amended virtual item group does not fit in one of the already existing item group clusters without creating any conflict. This sequence is in particular used when a new virtual item group is created, which is of course not jet classified.

[0030] In an embodiment, updating the classification may comprise, in case the amended virtual item group is already classified in one of the available item group clusters, to keep the amended virtual item group in the respective item group cluster if the newly added or removed virtual item does not create a conflict within the item group cluster. In this case, the amendment does not make any reclassification necessary because even the added or removed virtual item does not create a conflict within the respective cluster.

[0031] In an embodiment, updating the classification may comprise, in case the amended virtual item group is already classified in one of the available item group clusters, to remove the amended virtual item group from the respective item group cluster and to newly classify the amended virtual item group if the respective item group cluster remains in conflict if the virtual item group being in conflict with the amended virtual item group is removed from the respective item group cluster. The conflicting virtual item group is the virtual item group of another user, which creates in combination with the amended virtual item group the conflict (e.g. both virtual item groups comprising the same shirts of the same brand but of different sizes). According to this embodiment, a conflict is e.g. creates with a plurality of other virtual item groups of the same cluster.

[0032] In an embodiment, updating the classification may comprise, in case the amended virtual item group is already classified in one of the available item group clusters, to keep the amended virtual item group in the respective item group cluster and the virtual item group being in conflict with the amended virtual item group is removed from the respective item group cluster and newly classified if the respective item group cluster remains conflict free if the virtual item group being in conflict with the amended virtual item group is removed from the respective item group cluster. This sequence is e.g. used if only one virtual item group is in conflict with the amended virtual item group. The simple and efficient solution in this case is to remove the conflicting virtual item group from the cluster and to newly classify it.

[0033] In an embodiment, the computer implemented method is configured, in case the amended virtual item group is already classified in an item group cluster and kept in the respective item group cluster, to merge the item group cluster with another item group cluster if the newly added virtual item or removed virtual item creates a situation where the respective item group cluster and the item group cluster can be merged without creating a conflict within the new merged item group cluster. This embodiment may be in particular applicable in case the virtual item is removed from the virtual item group, which may create the situation that the respective cluster comprising the amended virtual item group may be merged with another cluster without creating a new conflict. Merging creates the advantage that the number of the searchable virtual item groups of one cluster becomes larger.

[0034] In an embodiment, computer-implemented method further comprises the step of accessing a request database, which stores former requests and the respective results. The request database may be a different database or may be implemented in the item database or the item group database. The request database stores the request, which have been made by the different users and further user information, in particular the respective virtual item group of the item category of the respective users. Further, the results of the respective requests are also stored. In case a stored former request having the same parameters is found in the request database the result of this request is used by the user associated property algorithm for determining the user associated desired property of the selected virtual item. The user associated property algorithm may access the request database and searches if a match can be found. A match may be determined in that the request is made for the same selectable virtual item and that the virtual item group of the access user making the request is identical or does highly correspond to the virtual item group of the user of the former stored request. If this is the case, it is checked if the former result may be used and displayed to the user. Such a check may comprise to verify if the former request is not to old (e.g. older than two months) or to verify if the virtual item group of the respective user has recently been amended drastically. This embodiment, may offer an advantageous fast and reliable option to determine the user associated desired property by using former equivalent or similar requests.

[0035] In a further embodiment, each virtual item of the respective virtual item group of the users is provided with a weighting, in particular wherein more recently added virtual items are provided with a higher weighting compared to previous added virtual items, and wherein the weighting is used by the user associated property algorithm for determining the user associated desired property. According to this embodiment, the determination of the user associated desired property is performed by using the weighting of the different virtual items. E.g. in case a plurality of directly matching virtual item groups comprise the selected virtual item, both are used for the determination of the user associated desired property, but the more recently added virtual item corresponding to the selected virtual item is weighted higher. Further, it is conceivable that specific virtual items, which have specific properties e.g. which are overrepresented or which have an advantageous fit etc. are provided with a higher weighting compared to other virtual items. This may further increase the desired accuracy of the determination.

[0036] In a further embodiment, a user profile comprising user information e.g. body dimensions etc. may be used by the user associated property algorithm for determining the user associated desired property.

[0037] In a further embodiment, the user associated property algorithm uses in addition or alternatively a neural network, which uses as input data the selected virtual item and the virtual item group of the access user, the neural network being trained, preferably constantly trained, using the item group database of all users, and the neural network provides as output data the user associated desired property. The neural network may run on a cloud-based server and receives requests from the user via a respective connection, wired or wireless, and accesses the respective databases for retrieving the required information as input data and for training of the neural network. Further, the neural network may send the output data; in particular, the determined user associated desired property to the access user. It is also conceivable that the neural network is used in addition to the previously mentioned embodiments for determining of the user associated desired property. In a further embodiment, the neural network may be used within one item group cluster for determining the user associated desired property.

[0038] In a further embodiment, different virtual item groups may comprise virtual items having relevant item properties of a different unit. E.g., some virtual item groups comprise shoes having the size stored in EU scale and other virtual item groups comprise shoes having the size stored in US size. In other to ensure the required comparability, the different item properties may be normalized to a selected unit, e.g. the EU size for shoes. This normalization may run in the backend, the user can select his desired unit and receives the determined user associated desired properties in his selected unit or scale.

[0039] In an embodiment, the computer-implemented method is configured for determining a user associated size as the user associated desired property of a physical clothing item, the computer- implemented method comprising the steps of:

[0040] Selecting, by the access user, a virtual clothing item, representing the physical clothing item from a selectable set of virtual clothing items stored in a clothing item database, for which the user associated size is to be determined.

[0041] Accessing, or providing access to, a wardrobe database as the item group database, which stores for the access user and the plurality of other users a respective virtual wardrobe, representing a physical wardrobe, the virtual wardrobes comprising at least one virtual clothing item of the same clothing category as the selected virtual clothing item, each virtual clothing item being stored with a plurality of respective relevant clothing item properties. Accessing, or providing access, to a database comprises that the computer implemented method itself access the respective database or that it sends a request to another entity, which accesses the respective databases and may sent the respective results back. determining the user associated size of the physical clothing item represented by the selected virtual clothing item, the user associated desired size being determined by the user associated property algorithm, which uses at least one relevant clothing item property from the virtual item group of at least one other user form the item group database.

[0042] According to this embodiment, the computer-implemented method is used for the specific purpose of determining the user associated size of a physical clothing item. Every embodiment, described above and I or hereinafter with respect to the general determination of the user associated desired property is of course also applicable to the specific embodiment.

[0043] According to a further aspect, a data processing device comprising a processor for carrying out the computer-implemented method as described above or hereinafter is specified. The data processing device may be e.g. a computer a handheld device, a wearable device, which provides the required functionalities, in particular which offers a human machine interface for interacting with the user, access possibilities to the data bases and a processor or electronic circuit for carrying out the required calculations in particular for implementing and running of the user associated property algorithm.

[0044] According to a further aspect, a digital platform implemented as a computer program is specified, which comprises instructions, which, when the program is executed by a processor, cause the processor to carry out the computer implemented method. The digital platform is e.g. hosted on a server, in particular, a cloud based server, and provides access to it via different channels e.g. via the internet, a website or an application running on a device. The digital platform may further have access or is directly connected to the databases and may have the required functionalities that the user associated property algorithm may run. According to a further aspect, a computer-readable medium comprising instructions is specified, which, when executed by a computer, cause the computer to carry out the computer implemented method as described above or hereinafter.

[0045] It is to be understood that both the foregoing general description and the following detailed description present embodiments, and are intended to provide an overview or framework for understanding the nature and character of the disclosure. The accompanying drawings are included to provide a further understanding, and are incorporated into and constitute a part of this specification. The drawings illustrate various embodiments, and together with the description serve to explain the principles and operation of the concepts disclosed.

[0046] BRIEF DESCRIPTION OF THE DRAWINGS

[0047] The herein described disclosure will be more fully understood from the detailed description given herein below and the accompanying drawings, which should not be considered limiting to the disclosure described in the appended claims. The drawings are showing:

[0048] Fig. 1 a first schematic illustration of a plurality of physical wardrobes of a user and the corresponding virtual wardrobe of the respective user according to a first embodiment;

[0049] Fig. 2 a second schematic illustration of a first sequence of determining an user associated desired property of a physical item according to a first embodiment;

[0050] Fig. 3 a third schematic illustration of a second sequence of determining an user associated desired property of a physical item according to a second embodiment;

[0051] Fig. 4 a fourth schematic illustration of a plurality of item group clusters according to a first embodiment; Fig. 5 shows a first flow diagram illustrating schematically a plurality of steps performed by at least one electronic device for determining an user associated desired property of a physical item according to a first embodiment;

[0052] Fig. 6 shows a second flow diagram illustrating schematically a plurality of steps performed by at least one electronic device for updating databases according to a first embodiment;

[0053] Fig. 7 shows a third flow diagram illustrating schematically a plurality of steps performed by a user associated property algorithm for determining an user associated desired property of a physical item according to a second embodiment;

[0054] Fig. 8 shows a fourth flow diagram illustrating schematically a plurality of steps performed by at least one electronic device and a user associated property algorithm for determining an user associated desired property of a physical item according to a third embodiment;

[0055] Fig. 9 shows a fifth flow diagram illustrating schematically a plurality of steps performed by at least one electronic device for updating a cluster classification according to a first embodiment;

[0056] Fig. 10 shows a sixth flow diagram illustrating schematically a plurality of steps performed by at least one electronic device and a user associated property algorithm for determining an user associated desired property of a physical item according to a fourth embodiment;

[0057] Fig. 11 shows a seventh flow diagram illustrating schematically a plurality of steps performed by at least one electronic device for updating a cluster classification according to a second embodiment. DETAILED DESCRIPTION OF THE DRAWINGS

[0058] Reference will now be made in detail to certain embodiments, examples of which are illustrated in the accompanying drawings, in which some, but not all features are shown. Indeed, embodiments disclosed herein may be embodied in many different forms and should not be construed as limited to the embodiments set forth herein; rather, these embodiments are provided so that this disclosure will satisfy applicable legal requirements. Whenever possible, like reference numbers will be used to refer to like components or parts.

[0059] Figure 1 shows a first schematic illustration of a plurality of physical item groups 1 10, in particular of a plurality of physical wardrobes 1 10’, and the corresponding virtual item group 1 1 1 , in particular virtual wardrobe 1 1 1 ’, of a user 100. It is visible that the user 100 comprises two physical wardrobes 1 10’, one for shoes as clothing category 1 12’ or item category 1 12 and the other for t- shirts as clothing category 1 12’ or item category 1 12. Each physical wardrobe 1 10’ comprises a plurality of physical items 102, in particular, physical clothing items 102’, which fit and which are used by the user 100. In other words, the physical wardrobe 1 10’ is the wardrobe, which is used by the user 100 in daily life. Figure 1 further shows a plurality of virtual wardrobes 1 1 1 ’, which represent the physical wardrobe 1 10’ of this user 100. In other words, the virtual wardrobes 1 1 1 ’ is the digital twin of at least a portion of the physical wardrobe 1 10’ of the user 100. Each of the virtual wardrobes 1 1 1 ’ comprises a plurality of virtual items 103’, in particular virtual clothing items 103’, which represent the respective physical clothing item 102’. Each virtual clothing item 103’ is stored with item properties 105, in particular clothing item properties 105’. The item properties 105 and the clothing item properties 105’ may be subdivided in two groups, a first group comprises the relevant clothing item properties 105’, which are used for determining of desired properties of other clothing items, such relevant clothing item properties 105’ are e.g. the brand, the size, the fit etc. The second group comprises the additional clothing item properties 105’, like color of the item etc.

[0060] The different clothing items 1 10’, 1 1 1 ’ are for example of different brands and have a different size, but still perfectly fit the user 100, this is not uncommon because different brands use different cuts and sizes. The user 100 may have shoes of a brand A having the size 10 and of a brand B having the size 9.5 and both shoes fit perfectly. In case the user 100 would like to buy another pair of shoes e.g. of another brand C, he or she does not know the perfect fitting size.

[0061] The figures and the description of the figures focus on the used case of determining a user associated size 104’ of a physical clothing item 102’, which has been selected by the user 100. Nevertheless, other used cases for determining a user associated desired property 104 of a physical item 102 is of course also conceivable.

[0062] Figure 2 shows a second schematic illustration of a first sequence of determining a user associated desired property 104 of a physical item 102 according to a first embodiment. The physical item 102 is for example a physical clothing item 102’, which a user would like to buy, preferably online and the user associated desired property 104 is the user associated size 104’, which would fit perfectly to the user with respect to the respective clothing item 102’. Figure 2 shows an access user 101 , which is the user 100 accessing via e.g. an application running on a handheld device, the system or digital platform 500 used for determining the user associated desired property 104.

[0063] Figure 2 shows the access user 100, which selects out of a set of selectable virtual items 103 a selected victual item 103x, in particular a selected virtual clothing item 103x’. This is e.g. done via an application running on the handheld device of the access user 100. The application accesses an item database 200, in particular a clothing item database 200’, which stores the selectable set of virtual items 103. The selected virtual clothing item 103x’ is in this embodiment a shoe from brand A (=relevant clothing item property 105’). The best fitting size 104’ of this shoe is the user associated desired property 104, which should be determined for the access user 101 . The user associated desired property 104 is determined by a user associated property algorithm 150, which runs e.g. at least partially on the handheld application of the access user 101 or a server, in particular a cloud based server or a combination thereof.

[0064] Figure 2 shows an item group database 300, in particular a wardrobe database 300’, which stores for the access user 101 and a plurality of other users 100 the respective virtual item groups 1 1 1 , in particular the respective virtual wardrobes 1 1 1 ’. The user associated property algorithm 150 accesses the wardrobe database 300’ and searches according to this embodiment for directly matching virtual item groups 120 in plurality of available virtual item groups 1 1 1 of the plurality of users 100. A directly matching virtual item group 120 of another user 100 is a virtual item group 1 11 , which comprises the selected virtual item 103x, e.g. the shoe of brand A, and which comprises another virtual item 103, which is also present in the virtual item group 1 1 1 of the access user 101 , e.g. a shoe of brand B having the size 10. Both relevant item properties 105 need to match e.g. brand and size in this example. In case such a directly matching virtual item group 120 is found, the user associated property algorithm 150 uses the respective item property 105 of the found selected virtual item 103x of the other user 100 and determines and shows this item property 105 as user associated desired property 104 to the access user 101 . In the example as shown in Figure 2, the user associated size 104’ would be the size 9 for the selected shoe of brand A. It is therefore possible to determine very accurately the user associated desired property 104, e.g. the size, of the selected virtual item 103x by user data from other users 100.

[0065] Figure 3 shows a third schematic illustration of a second sequence of determining the user associated desired property 104 of a physical item 102, e.g. the shoe of brand A, according to a second embodiment. This embodiment is e.g. used in case no directly matching virtual item group 120 is found, as explained with respect to Figure 2. Further, this embodiment may be used in addition to the embodiment as explained with respect to Figure 2.

[0066] Figure 3 also shows schematically the process of selecting the virtual item 103x from the selectable set of virtual items 103 stored in the item database 200. This embodiment in particular differs in that the selected virtual item 103x is searched in an item group cluster 140, in particular it is searched in the item group cluster 140 in which the virtual item group 1 1 1 of the access user 101 is categorized. According to this embodiment, the item group database 300 comprises a classification in different item group clusters 140 of the virtual item groups 1 1 1 of the different users 100. The cluster classification is determined in that each item group cluster 140 comprises only virtual item groups 1 1 1 , which are free of a conflict with respect to each other, wherein a conflict is determined in that two virtual items 103 of different virtual item groups 1 1 1 (of the same item group cluster 140), which have one relevant item property 105 in common have a varying other item property 105. If this is not the case, the respective item group cluster 140 is free of a conflict. The item group cluster 140 as shown in Figure 3 is for example free of a conflict. The item group cluster 140 comprises four virtual item groups 1 1 1 from the access user 101 and three additional users 100. The first virtual item group 1 1 1 and the second virtual item group 1 1 1 comprise both the same shoe as virtual item 103 from brand B in size 10, the second virtual item group 1 1 1 and the third virtual item group 1 1 1 share both the shoe from brand C in size 9 and the third and the fourth virtual item group 1 1 1 share both the shoe from brand D in size 8. A conflict would for example be if the shoe from brand D would be present in size 8 in the fourth virtual item group 1 1 1 and in size 8.5 in the third virtual item group 1 1 1 .

[0067] Figure 3 further schematically shows that the user associated property algorithm 150 searches in the item group cluster 140 for the selected virtual item 103x, which is in the present case found in the fourth virtual item group 1 1 1 . The user associated property algorithm 150 uses the respective item property 105 of the found selected virtual item 103x and determines and shows this item property 105 as the user associated desired property 104 to the access user 101 . In the example as shown in Figure 3, the user associated size 104’ would be the size 9 for the selected shoe of brand A.

[0068] Figure 4 shows a fourth schematic illustration of a plurality of item group clusters 140. Figure 4 shows advantageously that the item group cluster 140 comprise only virtual item groups 1 1 1 of one item group category 1 12, e.g. of shoes or shirts. It is therefore common that the different virtual item groups 1 1 1 of the user 100 are categorized in different clusters 140.

[0069] Figure 5 shows a first flow diagram illustrating schematically a plurality of steps performed by at least one electronic device, in particular its processor, for determining the user associated desired property 104 of the physical item 102 according to a first embodiment. In the following paragraphs, described with reference to Figure 5 is a possible sequence of steps for determining the user associated desired property 104.

[0070] In step S1 , the access user 101 , selects a virtual item 103 from a selectable set of virtual items 103 stored in the item database 200, thereby determining the selected virtual item 103, for which the user associated desired property 104 should be determined. The selecting process is e.g. performed via an application or a website running on a device from the access user 101 . In step S2, access to the item group database 300 is provided. In other words, the item group database 300, which comprises the virtual item groups 1 1 1 of other users 101 is provided such that the user associated property algorithm 150 may access the item group database 300 or may receive the respective necessary data for determining the user associated desired property 104 from the item group database 300.

[0071] In step S3, the user associated property algorithm 150, determines the user associated desired property 104 of the physical item 102 represented by the selected virtual item 103x, wherein the user associated desired property 104 is determined by the user associated property algorithm 150, which uses at least one relevant item property 105 from the virtual item group 1 1 1 of at least one other user 100 from the item group database 300. The user associated property algorithm 150 may further use at least one relevant item property 105 from the virtual item group 1 1 1 of the access user 101 .

[0072] Figure 6 shows a second flow diagram illustrating schematically a plurality of steps performed by at least one electronic device, in particular its processor, for updating the item database 200 and I or the item group database 300. In the following paragraphs, described with reference to Figure 6 is a possible sequence of steps for updating these databases. Both databases may be implemented as a single database.

[0073] In step A1 , a new virtual item 103 is added to the item database 200, which updates the item database 200, such that the newly added virtual item 103 becomes selectable by users 100. This step is e.g. done by retailers, by brand companies or by individuals having the respective permissions.

[0074] In step A2, the user 100, adds a new virtual item group 1 1 1 to the item group database 300. The user 100 creates via the respective application on his handheld or computer device a new virtual item group 1 1 1 e.g. of shoes or of another item group category 1 12. It is of course also conceivable to delete a virtual item group 1 1 1 . In step A3, the user 100, updates an existing of his virtual item groups 1 1 1 , by adding a new virtual item 103 or by deleting a virtual item 103, which updates the item group database 300. The physical wardrobe of users 100 may change over time, which makes it necessary to update the respective item group database 300.

[0075] Figure 7 shows a third flow diagram illustrating schematically a plurality of steps performed by the user associated property algorithm 150, in particular running on a processor, for determining the user associated desired property 104 according to a second embodiment. In the following paragraphs, described with reference to Figure 7 is a possible sequence of steps for determining the user associated desired property 104.

[0076] In step S3a, which is a sub-step of step S3, described with respect to Figure 5, the user associated property algorithm 150 searches in directly matching virtual item groups 1 1 1 of other users for the selected virtual item 103x. Directly matching virtual item groups 1 1 1 have already been described with respect to Figure 2.

[0077] In case the selected virtual item 103x is found in one directly matching virtual item group 1 1 1 , the step S3b is executed by the user associated property algorithm 150. In step S3b the user associated desired property 104 of the selected virtual item 103x for the access user 101 is determined S3b, by extracting the respective desired relevant item property 105 from the found selected virtual item 103x from the at least one directly matching virtual item group 120 of the respective other user 100. The result may further be shown to the access user 101 .

[0078] In case the selected virtual item 103x is not found in one directly matching virtual item group 1 1 1 , the user associated property algorithm 150 may use an alternative possibility for determining the user associated desired property 104 as outlined with respect to Figure 8.

[0079] Figure 8 shows a fourth flow diagram illustrating schematically a plurality of steps performed by the user associated property algorithm 150, in particular running on a processor of at least one electronic device, for determining the user associated desired property 104 according to a third embodiment. In the following paragraphs, described with reference to Figure 8 is a possible sequence of steps for determining the user associated desired property 104.

[0080] In step S2, which corresponds mainly to the step S2 as described with respect to figure 5, the item group database 300 is provided or access to the item group database 300 is provided. Figure 8 differs in that the item group database 300 further comprises a respective item group cluster classification for the virtual item groups 1 1 1 .

[0081] In step S3 of Figure 8, the user associated property algorithm 150 determines the user associated desired property by further using the selected virtual item 103x and the item group cluster classification, which will be explained below.

[0082] In step S3c, the user associated property algorithm 150, searches in the respective item group cluster 140, in which the virtual item group 1 1 1 of the access user 101 is classified, for the selected virtual item 103x.

[0083] In case the selected virtual item 103x is found in the respective virtual item group cluster 140, the user associated property algorithm 150 determines in step S3d the user associated desired property 104 of the selected virtual item 103x by extracting the respective desired relevant item property 105 from the found selected virtual item 103x in the item group cluster 140.

[0084] In case the selected virtual item 103x is not found in the item group cluster 140, the user associated property algorithm 150 may use an alternative possibility for determining the user associated desired property 104 as outlined with respect to Figure 9.

[0085] In the additional optional step S3e, the user associated property algorithm 150, may determine the at least one other user 100, in whose virtual item group 1 1 1 the selected virtual item 103x is found. Information from this user may further be used for improving the accuracy of determining the user associated desired property 104. In the additional optional step S3f , the user associated property algorithm 150, searches if at least one virtual item group 1 1 1 of a different item category 1 12 of the found at least one other user 100 is also classified in the same item group cluster 140 of this item category 1 12 of the access user 101 and the user associated property algorithm 150 retrieves relevant item properties 105 of matching virtual items 103 of this item group cluster 140.

[0086] In the additional optional step S3g, the user associated property algorithm 150, determines the user associated desired property of the selected item 104 by further using the retrieved relevant virtual item properties 105 of the matching virtual items 103 of this other item group cluster 140. This may further increases the accuracy of the determination of the user associated desired property 104.

[0087] Figure 9 shows a fifth flow diagram illustrating schematically a step performed by at least one electronic device, in particular an updating algorithm running on a processor of the device, for updating the item group cluster classification according to a first embodiment.

[0088] In step U1 , the classification of the virtual item group 1 1 1 in the available item group clusters 140, is updated, in particular by a respective updating algorithm, in case the respective virtual item group 1 1 1 is amended by adding a new virtual item 103 or removing one of the virtual items 103 from the virtual item group 1 1 1 . This updates the item group database 300, comprising the classification of the virtual item groups 1 1 1 in the respective item group clusters 140.

[0089] Figure 10 shows a sixth flow diagram illustrating schematically a plurality of steps performed by at least one electronic device, in particular its processor, and a user associated property algorithm, running preferably on the processors, for determining an user associated desired property of a physical item according to a fourth embodiment. In the following paragraphs, described with reference to Figure 10 is a possible sequence of steps for determining the user associated desired property 104. This embodiment differs from the previous embodiments, in particular in that it combines the presented sequences. The sequence starts with step S1 , in which the access user 101 , selects the virtual item 103, for which the user associated desired property 104 should be determined. Regarding this step, reference is made to the Figures 2, 3 and 5.

[0090] In step R1 , the user associated property algorithm 150, queries if a request having the same or almost the same parameters has been made bevor. The user associated property algorithm 150 may access a request database 400, which stores former requests from the same user 101 and from other users 100. In case the same request or a similar former request is found, the sequence continues in step R2. In case no request is found, the sequence continues in step R3.

[0091] In step R2, the user associated property algorithm 150, queries if the former result was correct and if the former result is applicable to the new request of the access user 101 . If yes, the result is shown to the access user 101 . If no, the sequence continues in step R3.

[0092] In step R3, the user associated property algorithm 150, queries in at least one directly matching virtual item groups 120 if the selected virtual item 103x can be found. Reference is e.g. made to Figure 7 and the respective description. If the selected virtual item 103x is found in a directly matching virtual item group 120, the respective relevant item property data is used for determining the user associated desired property 104 and shown to the access user 101 . If the selected virtual item 103x is not found in at least one of directly matching virtual item groups 120, the sequence continuous in step R4.

[0093] In step R4, the user associated property algorithm 150, searches in the item group cluster 140, in which the virtual item group 1 1 1 of the access user 1 1 1 is classified, for the selected virtual item 103x. In case the selected virtual item 103x is found in the respective item group cluster 140, the user associated desired property 104 of the selected virtual item 103x, is determined, by extracting the respective desired relevant item property 105 from the found selected virtual item 103x in the item group cluster 140. The result may further be presented to the access user 101 . Reference is made to Figure 8 and the respective description. In case the selected virtual item 103x is not found in the respective cluster 140, the sequence continuous in step R5. In step R5, the user associated property algorithm 150, queries if the selected virtual item 103x may be found in other virtual item groups 1 1 1 of other users 100 if relevant parameters of the virtual items 103 in the virtual item group 1 1 1 of the access user 101 and the other users may be changed. E.g. the size of the shoes is changed from 9 to 10. If the selected virtual item 103x may be found in other virtual item groups 1 1 1 , e.g. directly matching due to the changed parameters, the respective relevant item properties 105 are extracted and used. If this is the case the sequence continuous in step R6. If this is not the case the sequence ends, without determining the user associated desired property 104. Adding additional virtual items 103 by the access user may improve the situation.

[0094] In step R6, the user associated property algorithm 150, queries if the result can be transferred such that the determined user associated desired property 104 fits to the access user 101 . If yes, the determined user associated desired property 104 is transferred and shown to the user. If not, the sequence ends, without determining the user associated desired property 104.

[0095] Figure 11 shows a seventh flow diagram illustrating schematically a plurality of steps performed by at least one electronic device, in particular its processor, and a respective classification algorithm, running preferably on the processors, for updating the item group cluster classification according to a second embodiment. In the following paragraphs, described with reference to Figure 1 1 is a possible sequence of steps for updating the item group cluster 140.

[0096] The sequence starts, preferably automatically, when a virtual item group 1 1 1 is amended by the respective user 100. Reference is herewith made e.g. to Figure 6 and the respective description.

[0097] In step C1 , the classification algorithm, queries if the amended virtual item group 1 1 1 , is already classified in an item group cluster 140. This may be the case if the virtual item group 1 1 1 is newly created. If the amended virtual item group 1 1 1 is not yet classified, the sequence continuous in step C2, otherwise in step C3.

[0098] In step C2, the classification algorithm queries if the amended virtual item group 1 1 1 fits in any one of the already existing item group clusters 140 without creating any conflict. If yes, the amended virtual item group 1 1 1 is added C2b to the respective existing item group cluster 140. If no, the amended virtual item group 1 1 1 is added C2a to a newly created item group cluster 140.

[0099] In step C3, the classification algorithm queries if the amended virtual item group 1 1 1 creates a conflict, due to the amendment, within the item group cluster 140 in which the amended virtual item group 1 1 1 is already classified. If no, the sequence continues in step C4, if yes, the sequence continues in step C5.

[0100] In step C4, the classification algorithm queries if due to the amended virtual item group 1 1 1 a situation is created, in which the respective item group cluster 140 may be merged with another item group cluster 140. If yes, both or more item group clusters 140 are merged C4a. If no, the sequence ends without any change in the classification.

[0101] In step C5, the classification algorithm queries if the respective item group cluster 140 comprising the amended virtual item group 1 1 1 remains valid (conflict free) if the conflicting item group 1 1 1 is removed from the item group cluster 140. In case the respective item group cluster remains invalid, the amended virtual item group 1 1 1 is removed or deleted C5a from the respective item group cluster 140 and newly classified C5b. In case the respective item group cluster remains valid, by removing the conflicting virtual item group 1 1 1 , the conflicting virtual item group 1 1 1 is removed or deleted C5c from the respective item group cluster 140 and newly classified C5d. Afterwards the sequence ends.

[0102] LIST OF REFERENCE SIGNS

[0103] 100 User 200’ clothing item database

[0104] 101 Access User 300 item group database

[0105] 102 physical item 300’ wardrobe database

[0106] 102’ physical clothing item 30 400 request database

[0107] 103 virtual item 500 digital platform

[0108] 103’ virtual clothing item

[0109] 103x selected virtual item 51 Selecting

[0110] 103x’ selected virtual clothing item 52 Accessing

[0111] 104 user associated desired property 35 S3 Determining

[0112] 104’ user associated size S3a Searching

[0113] 105 relevant item property S3b Determining by extracting

[0114] 105’ clothing item property S3c Searching in respective item group

[0115] 106 weighting cluster

[0116] 110 physical item group 40 S3d Determining by extracting

[0117] 110’ physical wardrobe S3e Determining the other user

[0118] 111 virtual item group S3f Searching

[0119] 111 ’ virtual wardrobe S3g Determining

[0120] 112 item category S4 Determining

[0121] 112’ clothing category 45 S5 Searching

[0122] 120 directly matching virtual item group S6 Determining

[0123] 121 direct match A1-A3 Adding or Updating

[0124] 140 item group cluster U1 Updating the classification

[0125] 150 user associated property algorithm R1 -R6 Sequence of determining

[0126] 151 neural network 50 C1 -C5 Sequence of cluster update

[0127] 200 item database

Claims

PATENT CLAIMS1 . A computer-implemented method for determining a user associated desired property (104) of a physical item (102), the computer-implemented method comprising the steps of: a. selecting (S1 ), by an access user (102), a virtual item (103x), representing the physical item (102), from a selectable set of virtual items (103) stored in an item database (200), for which the user associated desired property (104) is to be determined; b. providing (S2) an item group database (300), which stores for the access user(101 ) and a plurality of other users (100) a respective virtual item group (1 1 1 ), representing a physical item group (1 10), the virtual item groups (1 1 1 ) comprising at least one virtual item (103) of the same item category (1 12) as the selected virtual item (103x), each virtual item (103) being stored with a plurality of respective relevant item properties (105); c. determining (S3) the user associated desired property (104) of the physical item(102) represented by the selected virtual item (103x), wherein the user associated desired property (104) is determined by a user associated property algorithm (150), which uses at least one relevant item property (105) from the virtual item group (1 1 1 ) of at least one other user (100) from the item group database (300).

2. The computer-implemented method according to claim 1 , wherein the user associated user associated property algorithm (150) comprises the steps of: a. searching (S3a) in directly matching virtual item groups (120) of the item group database (300) for the selected virtual item (103x), the searched directly matching virtual item groups (120) of the other users (100) comprise at least one othervirtual item (103), which is also comprised in the virtual item group (1 1 1 ) of the access user (101 ) having the same relevant item properties (105); b. in case the selected virtual item (103x) is found in at least one directly matching virtual item group (120) the user associated desired property (104) of the selected virtual item (103x) for the access user (101 ) is determined (S3b), by extracting the respective desired relevant item property (105) from the found selected virtual item (103x) from the at least one directly matching virtual item group (120) of the respective other user (100).

3. The computer-implemented method according to claim 2, wherein the user associated property algorithm (150) is further configured to determine the user associated desired property (104) of the selected virtual item (103) in case the selected virtual item (103x) is found in a plurality of virtual item groups (1 1 1 ), preferably directly matching virtual item groups (120), by extracting and using the desired relevant item property (105) of the found selected virtual item (103x) from the plurality of virtual item groups (1 1 1 ) of other users (100).

4. The computer-implemented method according to any one of the preceding claims, further comprising the steps of: a. adding (A1 ) a new virtual item (103) to the item database (200), thereby updating the item database (200), the newly added virtual item (103) becomes thereby selectable by at least one user (100); and I or b. adding (A2) a new virtual item group (1 1 1 ) of a new or existing user (100) and I or updating (A3) an existing virtual item group (1 1 1 ) by adding a new virtual item (103) or by deleting at least one virtual item (103), thereby updating the item group database (300).

5. The computer-implemented method according to any one of the preceding claims, wherein, the computer-implemented method is, in particular in case the selected virtual item (103x) is not found in at least one directly matching virtual item group (120), configured to: a. providing (S2) the item group database (300), in which each virtual item group (1 1 1 ) is classified in an item group cluster (140) based on the virtual items (103) of the respective virtual item group (1 1 1 ) and the respective relevant item properties (105) of the virtual items (103), wherein each item group cluster (140) comprises only virtual item groups (1 1 1 ), which are free of a conflict with respect to each other, wherein a conflict is determined in that two virtual items (103) of different virtual item groups (1 1 1 ) which have one relevant item property (105) in common have a varying other item property (105); and the user associated property algorithm (150) further comprises the steps of: b. searching (S3c) in the item group cluster (140), in which the virtual item group (1 1 1 ) of the access user (1 1 1 ) is classified, for the selected virtual item (103x); c. in case the selected virtual item (103x) is found in the item group cluster (140), in which the virtual item group (1 1 1 ) of the access user (1 1 1 ) is classified, the user associated desired property (104) of the selected virtual item (103x), is determined (S3d), by extracting the respective desired relevant item property (105) from the found selected virtual item (103x) in the item group cluster (140).

6. The computer-implemented method according to claim 5, wherein in case the selected virtual item (103x) is found in the item group cluster (140), in which the virtual item group (1 1 1 ) of the access user (101 ) is classified, the user associated property algorithm (150) further comprises the steps of:a. determining (S3e) the at least one other user (100) in whose virtual item group(1 1 1 ) the selected virtual item (103x) is found; and b. searching (S3f) if at least one virtual item group (1 1 1 ) of a different item category(1 12) of the at least one other user (100) is also classified in the same item group cluster (140) of this item category (1 12) of the access user (101 ) and retrieving relevant item properties (105) of matching virtual items (103) of this item group cluster (140); c. determining (S3g) the user associated desired property of the selected item (104) by further using the retrieved relevant virtual item properties (105) of the matching virtual items (103) of this item group cluster (140).

7. The computer-implemented method according to any one of the claims 5 or 6, further comprising the step: a. updating (U1 ) the classification of the virtual item group (1 1 1 ) in the available item group clusters (140), in case the respective virtual item group (1 1 1 ) is amended by adding a new virtual item (103) or removing a virtual item (103) from the virtual item group (1 1 1 ).

8. The computer-implemented method according to claim 7, wherein in case the amended virtual item group (1 1 1 ) is not jet classified in any of the available item group clusters (140), it is added (C2b) to an already existing item group cluster (140) if the amended virtual item group (1 1 1 ) fits in one of the already existing item group clusters (140) without creating any conflict, or it is added (C2a) to a newly created item group cluster (140) if the amended virtual item group (1 1 1 ) does not fit in one of the already existing item group clusters (140) without creating any conflict.

9. The computer-implemented method according to one of the claims 7 or 8, wherein in case the amended virtual item group (1 1 1 ) is already classified in one of the available item groupclusters (140), the amended virtual item group (1 1 1 ) is kept in the respective item group cluster (140) if the newly added or removed virtual item (103) does not create a conflict within the item group cluster (140), or the amended virtual item group (1 1 1 ) is removed (C5a) from the respective item group cluster (140) and newly classified (C5b) if the respective item group cluster (140) remains in conflict if the virtual item group (1 1 1 ) being in conflict with the amended virtual item group (1 1 1 ) is removed from the respective item group cluster (140), or the amended virtual item group (1 1 1 ) is kept in the respective item group cluster (140) and the virtual item group (1 1 1 ) being in conflict with the amended virtual item group (1 1 1 ) is removed (C5c) from the respective item group cluster (140) and newly classified (C5d) if the respective item group cluster (140) remains conflict free if the virtual item group (1 1 1 ) being in conflict with the amended virtual item group (1 1 1 ) is removed from the respective item group cluster (140).

10. The computer-implemented method according to claim 9, wherein in case the amended virtual item group (1 1 1 ) is already classified in an item group cluster (140) and kept in the respective item group cluster (140), the item group cluster (140) is merged (C4a) with another item group cluster (140) if the newly added virtual item (103) or removed virtual item (103) creates a situation where the respective item group cluster (140) and another one of the item group cluster (140) can be merged without creating a conflict within the new merged item group cluster (140).1 1 . The computer-implemented method according to any one of the preceding claims, further comprising the steps of: a. accessing (R1 ) a request database (400), which stores former requests and the respective results;b. in case a stored former request having the same or similar parameters is found in the request database (400) the result of this request is used by the user associated property algorithm (150) for determining the user associated desired property (104) of the selected virtual item (103x).

12. The computer-implemented method according to any one of the preceding claims, wherein each virtual item (103) of the respective virtual item group (1 1 1 ) of the users (100) is provided with a weighting (106), in particular wherein more recently added virtual items (103) are provided with a higher weighting (106) compared to previous added virtual items (103), and wherein the weighting (106) is used by the user associated property algorithm (150) for determining the user associated desired property (104).

13. The computer-implemented method according to any one of the preceding claims, wherein the user associated property algorithm (150) uses a neural network (151 ), which uses as input data the selected virtual item (103x) and the virtual item group (1 1 1 ) of the access user (101 ), the neural network (151 ) being trained, preferably constantly trained, using the item group database (300) of all users (100), and the neural network (151 ) provides as output data the user associated desired property (104).

14. The computer-implemented method according to any one of the preceding claims, being configured for determining a user associated size (104’) as the user associated desired property (104) of a physical clothing item (102’), the computer-implemented method comprising the steps of: a. selecting (S1 ), by the access user (102), a virtual clothing item (103x’), representing the physical clothing item (102’) from a selectable set of virtual clothing items (103’) stored in a clothing item database (200’), for which the user associated size (104’) is to be determined; b. providing (S2) a wardrobe database (300’) as the item group database (300), which stores for the access user (101 ) and the plurality of other users (100) arespective virtual wardrobe (1 1 1 ’), representing a physical wardrobe (1 1 1 ), the virtual wardrobes (1 1 1 ) comprising at least one virtual clothing item (103’) of the same clothing category (1 12’) as the selected virtual clothing item (103x’), each virtual clothing item (103’) being stored with a plurality of respective relevant clothing item properties (105’); c. determining (S3) the user associated size (104’) of the physical clothing item (102’) represented by the selected virtual clothing item (103x’), the user associated desired size (104’) being determined by the user associated property algorithm (150), which uses at least one relevant clothing item property (105’) from a virtual wardrobe (1 1 1 ’) of at least one other user (100) from the wardrobe database (300’).

15. A data processing device comprising a processor for carrying out the computer-implemented method of any one of the claims 1 to 14.

16. A digital platform (500) implemented as a computer program comprising instructions, which, when the program is executed by a processor, cause the processor to carry out the computer implemented method of any one of claims 1 -14.

17. A computer-readable medium comprising instructions, which, when executed by a computer, cause the computer to carry out the computer-implemented method of any one of claims 1 -14.

Images