DESC:FIELD OF THE INVENTION:
This invention relates to the field of monitoring devices, computer engineering, nutrition science, data analytics, and neural networks.

Particularly, this invention relates to a system and method for providing personalised dietary recommendations based on multiple parameters.

BACKGROUND OF THE INVENTION:
There is an increasing consciousness towards improving diet and lifestyle to avoid the incidence of or managing of several diseases, specifically obesity, cardiovascular issues, and diabetes. The increasing scourge of lifestyle diseases begs interventional directed steps towards improving a typical diet, which often comes with a price of either compromising on taste and / or satiety. Typical diet interventions, especially for weight loss, focus on restricting calorific intake.

Nutritionists, dieticians, and doctors advocate the intake of food items generally perceived to be healthy, i.e. raw fruits and vegetables, and/or avoiding specific food group like fat, carbohydrates etc.

While there isn’t a single plan that has worked for all users, even the relatively effective plans are based on restrictive measures and donot incorporate user behaviour and the nature of metabolic systems in a human body that are responsible for quick regain of weight as soon as some of these restrictions are relaxed. Moreover, these plans are not repeatable or objectively defined.

Furthermore, this science, at best has been vastly subjective in nature because of lack of repeatability across various conditions. These various conditions comprise characteristics of a human body, which differ across individuals.

Therefore, there is a need for a system and method which converts these subjective parameters into an objective science using technology to parameterise the subjective items and to define rules for correlating and mapping these subjective items to attain repeatable objective recommendations that work at each individual level.

Typically, this system and method is configured to provide a user with dietary recommendations which map on to a plurality of parameters comprising fullness parameters, taste parameters, and health parameters.

OBJECTS OF THE INVENTION:
An object of the invention is to provide a system and method for providing personalised diet recommendations that incorporate defined multiplicity of quotients relating to food and body.

Another object of the invention is to provide a system and method for providing personalised diet recommendations that incorporate defined multiplicity of quotients relating to food and body, said one of the multiplicity of quotients being a satiety quotient.

Yet another object of the invention is to provide a system and method for providing personalised diet recommendations that incorporate defined multiplicity of quotients relating to food and body, said one of the multiplicity of quotients being a health quotient

Still another object of the invention is to provide a system and method for providing personalised diet recommendations that incorporate defined multiplicity of quotients relating to food and body, said one of the multiplicity of quotients being a taste quotient

An additional object of the invention is to parameterise a satiety quotient in terms of content items pertaining to a human body based on pre-defined parameters of a human body in order to provide dietary recommendations which correlate with and conform to content items of the satiety quotient as well as the pre-defined parameters of the human body.

Yet an additional object of the invention is to parameterise a health quotient in terms of content items pertaining to a human body based on pre-defined parameters of a human body in order to provide dietary recommendations which correlate with and conform to content items of the health quotient as well as the pre-defined parameters of the human body

Still an additional object of the invention is to parameterise a fullness quotient in terms of content items pertaining to a human body based on pre-defined parameters of a human body in order to provide dietary recommendations which correlate with and conform to content items of the taste quotient as well as the pre-defined parameters of the human body.

SUMMARY OF THE INVENTION:
According to this invention, there is provided a system and method for a system providing personalised dietary recommendations based on multiple parameters, said system comprises:
a first storage device networked within said system, said first storage device configured to store content items pertaining to a food item;
a second storage device networked within said system, said second storage device configured to store content items pertaining to calorific value of food items of said first storage device;
a third storage device networked within said system, said third storage device configured to store content items pertaining to geographic location of food items of the first storage device;
an attribute manager configured to determine and store attribute-related content items pertaining to food items stored in said first storage device;
a first selection mechanism configured to prompt a user to select at least a choice of food item pertinent to said user;
a second selection mechanism configured to prompt a user to select at least a food item that a user has ingested along with time, date, and serving size;
an input mechanism configured to prompt a user to input said user’s details relating to at least one of height data, weight data, age data, gender data, location data, ethnicity data, genomic data, genetic data, and pertinent body data, characterised in that, said input mechanism comprising an ecosystem of wearables configured to provide distributed nodes as input mechanisms;
a metabolic profiling mechanism is configured to read and store a metabolic profile of a user;
a first dynamic GUI generator configured to generate a user-specific dynamic GUI corresponding a user’s first synthesized profile correlative to said user’s taste quotient, said first dynamic GUI generator being communicably coupled to said first selection mechanism, said second selection mechanism, said input mechanism, and said metabolic profiling mechanism;
a second dynamic GUI generator configured to generate a user-specific dynamic GUI corresponding a user’s second synthesized profile correlative to said user’s satiety quotient, said second dynamic GUI generator being communicably coupled to said first selection mechanism, said second selection mechanism, said input mechanism, and said metabolic profiling mechanism;
a third dynamic GUI generator configured to generate a user-specific dynamic GUI corresponding a user’s third synthesized profile correlative to said user’s health quotient, said third dynamic GUI generator being communicably coupled to said first selection mechanism, said second selection mechanism, said input mechanism, and said metabolic profiling mechanism;
a taste parameter configuration engine configured to create a taste profile for each food item, characterised in that, each food item being defined by taste vectors and stored in a taste-related storage device, each of said taste vectors being correlated to a food item;
a first correlation engine, governed by a first rule engine, correlating a first synthesised profile of said user with a taste profile of a food item to determine a score of relevancy of taste for said user with respect to said food item;
a satiety parameter configuration engine configured to create a satiety profile for each food item, characterised in that, each food item being defined by satiety vectors and stored in a satiety-related storage device, each of said satiety vectors being correlated to a food item;
a second correlation engine, governed by a second rule engine, correlating a second synthesised profile of said user with a satiety profile of a food item to determine a score of relevancy of satiety for said user with respect to said food item;
a health parameter configuration engine configured to create a health profile for each food item, characterised in that, each food item being defined by health vectors and stored in a health-related storage device, each of said health vectors being correlated to a food item;
a third correlation engine, governed by a third rule engine, correlating a third synthesised profile of said user with a health profile of a food item to determine a score of relevancy of health for said user with respect to said food item; and
a recommendation engine configured to provide a food item output from the first storage device based on outputs from said first correlation engine said second correlation engine, and said third correlation engine.

Typically, said system comprising a calorific computation engine configured to receive a user’s weight goal in order to determine at least an appropriate cut in the calorific intake of the user and provide said data to said first correlation engine, said second correlation engine, and said third correlation engine in order to compute a staggered time-defined, goal-defined calorific data per user per time period per goal.

Typically, said system comprising a meal evaluation engine configured to receive data from said second selection mechanism in order to output data relating to a user’s ingested meal into pre-defined attribute content items.

Typically, said system comprising an activity monitoring module from said ecosystem of wearables configured to monitor physical activity of a user and store it in terms of activity data items.

Typically, said system comprising a taste vector mapping engine configured to map said taste vectors for a taste profile for a food item in order to provide relevant recommendations from said recommendation engine.

Typically, said system comprising a taste vector mapping engine configured to map said taste vectors for a taste profile for a food item in order to provide relevant recommendations from said recommendation engine, characterized in that, said taste vector mapping engine being configured to map each food item in a six-dimensional space array, each array correlative to at least a taste type and each item of each array of each taste type being populated with an intensity signal for that taste type.

Typically, said system comprising a taste vector mapping engine configured to map said taste vectors for a taste profile for a food item in order to provide relevant recommendations from said recommendation engine, characterized in that, said taste vector mapping engine being configured to map each food item in a six-dimensional space array, each array correlative to at least a taste type and each item of each array of each taste type being populated with an intensity signal for that taste type, further characterized in that, said taste vectors being weighted and normalised taste vectors based on ingredients per food item.

Typically, said system comprising a satiety vector mapping engine configured to map said satiety vectors for a satiety profile for a food item in order to provide relevant recommendations from said recommendation engine.

Typically, said system comprising a satiety vector mapping engine configured to map said satiety vectors for a satiety profile for a food item in order to provide relevant recommendations from said recommendation engine, characterized in that, said space vector mapping engine being configured to map each food item with a user-specific calorific target based on data from a group consisting of food ingredient composition and time of ingestion.

Typically, said system comprising a satiety vector mapping engine configured to map said satiety vectors for a satiety profile for a food item in order to provide relevant recommendations from said recommendation engine, characterized in that, said space vector mapping engine being configured to map each food item with a user-specific calorific target based on data from a group consisting of food ingredient composition and time of ingestion, further characterized in that, said satiety quotient being correlative to a body weight and body type.

Typically, said system comprising a health vector mapping engine configured to map said health vectors for a health profile for a food item in order to provide relevant recommendations from said recommendation engine.

Typically, said first storage device being a set of relationally-defined interconnected devices¸ wherein the set of devices cumulatively comprises at least a content item pertaining to an identity of a food item per se, at least a content item relating to recipe of the food items, at least a content item relating to ingredients of the food item, at least a content item relation to at least a nutrient of the food item, at least a content item relating to a pre-defined parameter of the food item.

Typically, said second storage device being a set of relationally-defined interconnected devices¸ wherein the set of devices cumulatively comprises at least a content item pertaining to calorific values of food items as well as ingredients of the food item.

Typically, said third storage device being a set of relationally-defined interconnected devices¸ wherein the set of devices cumulatively comprises at least a content item pertaining to geographic location(s) of the food items, ingredients of the food item, cultural attributes of the food items.

Typically, said input mechanism being configured to receive user inputs, wherein each input being correlated to a signal comprising a content item fetched from a group of storage devices comprising said first storage device, said second storage device, said third storage device, along with a signal comprising data from said attribute manager, and a signal comprising data correlating to a time parameter.

Typically, said recommendation engine is governed by a rule engine, said rule engine receiving inputs from said first storage device, said second storage device, said third storage device, said first selection mechanism, said second selection mechanism, said input mechanism, said metabolic profiling mechanism, said calorific computation engine, said meal evaluation engine, said activity monitoring module, said taste parameter configuration engine, said first correlation engine, said satiety parameter configuration engine, said second correlation engine, said health parameter configuration engine, said third correlation engine, in order to output a signal signal which comprises at least a content item correlating to a food item, said output signal having a cumulative strength of a signal corresponding to a health quotient, a signal corresponding to a taste quotient, and a signal corresponding to a satiety quotient.

BRIEF DESCRIPTION OF THE ACCOMPANYING DRAWINGS:
The invention will now be described in relation to the accompanying drawings, in which:
Figure 1 illustrates a schematic block diagram of the system of this invention;
Figure 2 illustrates a flowchart depicting steps of obtaining a taste quotient;
Figure 3 illustrates a flowchart depicting steps of a pairing method; and
Figure 4 illustrates a flowchart depicting steps of a providing a content item relating to a food item recommendation.

DETAILED DESCRIPTION OF THE ACCOMPANYING DRAWINGS:
According to this invention, there is provided a system and method for a system and method for providing personalised dietary recommendations based on multiple parameters.

Figure 1 illustrates a schematic block diagram of the system of this invention.

In accordance with an embodiment of this invention, a first storage device (D1) is networked with other elements, devices, and mechanisms of this system. The first storage deviceis configured to store content items pertaining to food. Typically, this device is a set of relationally-defined interconnected devices which comprises at least an identity of a food item per se, at least a content item relating to recipe of the food items, at least a content item relating to ingredients of the food item, at least a content item relation to at least a nutrient of the food item, at least a content item relating to a pre-defined parameter of the food item. Furthermore, content items may pertain to preparation and cooking time, seasonal information (relevant to fresh fruits and certain recipes), geographical preferences, food source (homemade, processed, etc), allergen content, food group and identity (vegetarian, grains, proteins, etc.)

In at least an embodiment, this first storage device comprises information of typical macronutrients and the calorific content of each of the food items, i.e. fat, protein, and carbohydrates. Important components in context of weight loss, sodium, and sugar will also be collected and stored. Additionally, the firststorage devicecomprises key micronutrients like vitamins and minerals. This information is collected for each ingredient, thereby enabling the estimation of any other food item based on its recipe.

In accordance with an embodiment of this invention, a second storage device (D2) is networked with other elements, devices, and mechanisms of this system.Typically, this device is a set of relationally-defined interconnected devices whichis configured to store content items pertaining to calorific value of food items of the first storage device.Each food item is tagged with pertinent calorific values of the food items per se as well as ingredients of the food item per se.

In accordance with an embodiment of this invention, a third storage device (D3) is networked with other elements, devices, and mechanisms of this system.Typically, this device is a set of relationally-defined interconnected devices which is configured to store content items pertaining to geographic location of food items of the first storage device. Each food item is tagged with pertinent geographic location(s) of the food items per se as well as ingredients of the food item per se. Furthermore, this second storage device may comprise content items pertaining to cultural attributes of the food items of the first storage device. Each food item is tagged with pertinent cultural attribute(s) of the food items per se as well as ingredients of the food items per se.

In accordance with an embodiment of this invention, an attribute manager module is configured to determine and store attributes pertaining to food items stored in the first storage device.

In accordance with an embodiment of this invention, a first selection mechanism (SM1) is configured to prompt a user to select at least a choice of food item pertinent to the user. Since each food item is tagged with attributes forming pertinent content items, these attributes are stored in a co-relational manner with respect to a user for use by this system and method. These selected food items are used to retrieve taste attribute content items of a user in order to map it to recommendations provided to a user.

A user’s profile is synthesised into a first dataset of content items wherein this first synthesised profile comprises content items correlative to a user’s taste quotient. This first synthesised profile is synthesised by means of a first dynamic GUI generated by a GUI generator wherein a user-specific dynamic GUI is formed to provide a single synthesised view of the user. Each input of the user correlates to a signal comprising a content item fetched from a group of storage devices comprising the first storage device, the second storage device, the third storage device, along with a signal comprising data from attribute manager, and a signal comprising data correlating to a time parameter.

Figure 2 illustrates a flowchart depicting steps of obtaining a taste quotient.

A user inputs data relating to a food item (202) which is considered and from a relevant data storage device (204), relevant content items (206) are obtained. These content items comprise cuisine, food time, class ingredients and their proportions, ingredient properties (odour, taste), cooking style and the like. A user may input feedback (208) regarding recommendations. From such feedback, relevant content items such as important ingredients and estimated taste of final food item is derived (210). Based on this data, another storage device (214) is used to output data relating to a food item (216). A mechanism is configured to correlate and determine (218) similar food items by ingredients as also distributions (220) based on cuisine, class of food item, and the like. Further, a similarity score (222) is established based on ingredient intersection similarity score based on cuisine closeness similarity score based on class closeness. All this data is used to determine a cumulative weighted similarity (224) which is further used to determine an output of a content item which correlated with a corresponding food item recommendation from a data storage device (226).

In accordance with an embodiment of this invention, a second selection mechanism (SM2) is configured to prompt a user to select at least a food item that a user has ingested along with time and date and serving size. This enables the system and method to track intake and patterns. It stores and records the user’s food consumption history. Since each food item is tagged with attributes forming pertinent content items, these attributes are stored in a co-relational manner with respect to a user for use by this system and method. These selected food items are used to retrieve satiety attribute content items of a user in order to map it to recommendations provided to a user.A user’s profile is synthesised into a second dataset of content items wherein this second synthesised profil ecomprises content items correlative to a user’s satiety quotient. Furthermore, these selected food items are used to retrieve health attribute content items of a user in order to map it to recommendations provided to a user.A user’s profile is synthesised into a third dataset of content items wherein this third synthesised profilecomprises content items correlative to a user’s health quotient.

A user’s profile is synthesised into a second dataset of content items wherein this second synthesised profile comprises content items correlative to a user’s satiety quotient. This second synthesised profile is synthesised by means of a second dynamic GUI generated by a GUI generator wherein a user-specific dynamic GUI is formed to provide a single synthesised view of the user. Each input of the user correlates to a signal comprising a content item fetched from a group of storage devices comprising the first storage device, the second storage device, the third storage device, along with a signal comprising data from attribute manager, and a signal comprising data correlating to a time parameter.

A user’s profile is synthesised into a first dataset of content items wherein this third synthesised profile comprises content items correlative to a user’s health quotient. This third synthesised profile is synthesised by means of a third dynamic GUI generated by a GUI generator wherein a user-specific dynamic GUI is formed to provide a single synthesised view of the user. Each input of the user correlates to a signal comprising a content item fetched from a group of storage devices comprising the first storage device, the second storage device, the third storage device, along with a signal comprising data from attribute manager, and a signal comprising data correlating to a time parameter.

In accordance with an embodiment of this invention, an input mechanism (IM) is configured to prompt a user to input a user’s details relating to at least one of height data, weight data, age data, gender data, location data, ethnicity data, genomic data, genetic data, and the like pertinent body data.

Figure 3 illustrates a flowchart depicting steps of a pairing method.

A user inputs data relating to a food item (302) which is considered. Using this input, the food item is mapped (304) as a node in a graph. If the specific content item relating to a food item is logged (306) more than a pre-defined number of times, then its position node is correlated (308) with most similar pairings in terms of taste and class. If the specific content item relating to a food item is not logged (306) more than a pre-defined number of times, then its position node is based (310) on terms of class information and taste profile. In either instance, edge weights of the content item are calculated (312) with log frequency and class log frequency node position. Further, directional edges are assigned (314) in the graph. This is stored (316) in a data storage device (318).

In accordance with an embodiment of this invention, a metabolic profiling mechanism is configured to read and store a metabolic profile of a user. This user metabolic profiling mechanism allows the system and method of this invention to enable a user to arrive at user’s calorific targets for individual meals and desired weight gain/loss. In at least one embodiment, a user metabolic profile is communicably coupled with user-defined calorific targets or system-defined calorific targets in order to provide pertinent recommendations.

An ecosystem of wearables and other such input mechanism may be configured to provide distributed nodes as input mechanisms for recording intake. Metabolic profiles may be captured through such input mechanisms.

Since each food item is tagged with attributes forming pertinent content items, these attributes are stored in a co-relational manner with respect to a user for use by this system and method. These selected food items are used to retrieve metabolic attribute content items of a user in order to map it to recommendations provided to a user.

In at least an embodiment, the system and method of this invention uses a linearized form of the model in its closed form analytical solution to arrive at a typical calorific requirement for the user based on input data items. A calorific computation engine (CCE) is configured to receive a user’s weight goal in order to determine at least an appropriate cut in the calorific intake of the user.

In at least an embodiment, the calorific computation engineis configured to compute a staggered time-defined, goal-defined calorific data per user per time period per goal. Feedback over successive time intervals and user inputs is used to re-define the calorific computation engine.

In accordance with an embodiment of this invention, a meal evaluation engine (MEE) configured to receive data from the second selection mechanism in order to output data relating to a user’s ingested meal into pre-defined attribute content items. Typically, this meal evaluation engine evaluated a user’s current meal in order to suggest modifications to the user’s meals. Data from the first storage device, the second storage device, and the third storage device is used.

In accordance with an embodiment of this invention, an activity monitoring module from the ecosystem of wearables and other such input mechanisms is configured to monitor physical activity of a user and store it in terms of activity data items. This enables the system and method is used to understand the energy expenditure of the user.

In accordance with an embodiment of this invention, a taste parameter configuration engine (TE) is configured to create a taste profile for each food item. A taste vector mapping engine is configured to map said taste vectors for a taste profile for a food item in order to provide relevant recommendations from said recommendation engine.Each food item is defined by taste vectors and stored in a taste storage device, each of the taste vectors being correlated to a food item. A plurality of taste vectors for a taste profile for a food item. These taste vectors and correlations are further used in order to map it to recommendations provided to a user.

A first correlation engine (CE1), governed by a first rule engine, correlates a first synthesised profileof a user with a taste profile of a food item to determine a score of relevancy of taste for a user with respect to the food item. This taste score is a component of a taste quotient whilst recommending a food item.

The food and user taste profile is generated and stored as a vector record in six dimensional taste space referred to as the taste space. Each dimension of this space is an elementary taste ‘Saltiness’, ‘Sweetness’, ‘Sourness’, ‘Saltiness’, ‘Bitterness’, ‘Umami’, and ‘Hot’. The value associated with each of these dimension is assumed to lie between 0 and 1, 1 signifying the maximal intensity and 0 being the minimum. The hypercube enclosed by each of this dimension is the taste space, and any individual food item lies within the hypercube. This hypercube then becomes the universal domain within which all food items lie.

Each ‘ingredient’ in a recipe is pre-assigned a level of flavour/ taste in each of the six dimensions. This number is assigned based on our expertise and experience. For example, the ingredient salt has a value 1 in the dimension of ‘saltiness’ while having a value 0 in all other dimensions. Lemon Juice is assumed to have 0 in all dimensions, but a value of 1 in ‘sourness’. Water can be assumed to have a value 0 in all dimensions.

Next, for any other food item, the ingredients taste vectors are now added weighted on the relative contribution in the recipe and the composite taste vector for the food item is computed. Finally it is normalized so that each value in the vector (i.e. contribution in each taste dimension) lies between 0 and 1. For example for lemonade, with ingredients as water (250 grams), lemon juice (20 grams), sugar (10 grams), and salt (2.5 grams). The composite taste vector would be 20*1 +0*250 +0*10+ 0*2.5 = 20/282.5 (= 0.07) in sourness, and similarly 10/282.5 (=0.035) in sweetness, and 2.5/282.5 (=0.009) in saltiness. It will have 0 in the ‘Hot’ dimension, since it does not have any ingredient that contributes to that dimension.

Additionally the food item is characterized by two more attributes, the texture and the smell. Those additional attributes are also collected and stored with discrete levels and used to characterize the food item.
Based on above mechanism, the system and method of this invention can compute the taste vector for any food item in the storage device and also characterize it in form of texture and smell.

For generation of a user taste profile, a user logs frequently eaten data and liked food data is collected. Each of these items, in the taste dimension constitutes what is hereby called user taste preference. The unique features of these points in the taste space are represented by the location of individual points. For large datasets, wherein the user log has several items, the features are extracted using hierarchical clustering algorithms, to enable speedy computations of subsequent food item likelihoods.

In accordance with an embodiment of this invention, a satiety parameter configuration engine (SE) is configured to create a satiety profile for each food item. A satiety vector mapping engine is configured to map said satiety vectors for a satiety profile for a food item in order to provide relevant recommendations from said recommendation engine.Each food item is defined by satiety vectors and stored in a satiety storage device, each of the satiety vectors being correlated to a food item. A plurality of satiety vectors for a satiety profile for a food item. These satiety vectors and correlations are further used in order to map it to recommendations provided to a user.

The satiety parameter configuration engine invokes a function of mathematical optimisation for meal size. Each food item in a meal that is to be recommended is further optimised based on user calorific target. This is accomplished by adding individual calorific values of the content items pertaining to a food items in a meal and scaling them to meet an overall meal target in a simultaneous optimisation. In at least one embodiment, the targets are set based on individual macronutrient composition bases (fat, protein, carbohydrates) and also for each meal time (breakfast, lunch, snack, dinner).

A feedback mechanism allows the system and method to be self-learning, to recalibrate, to re-estimate, and to refine a user’s taste and metabolic profiles in order to provide better recommendations based on a user’s objective.

Food preparation and recipes have a lot of intrinsic variability. The same food item, depending on the preparation can have widely different nutritional content depending upon the amount and quality of the ingredient used. Additionally, the metric of serving size also varies with the user. In order to account for this, the system and method of this invention refines the food intake term in a metabolic profile model for an individual user by introducing an adjustment factor that can account for this variability in broad terms without going into the tedious details of the specific user recipe and accurately measuring his food servings. The system and method hypothesizes that in the course of time with large enough data, this factor will converge to a user specific value that can be considered a constant.

By gathering the time course of user weight data and estimating the efficiency factor of food intake, this can be achieved as follows:

The system and method of this invention is governed, in part, by the following mathematical model as the basis of describing an individual’s body weight as a function of the calorific intake. Body weight typically follows the following equation:

dBW/dt+a(BW-BW_0 )=(?_i¦(1/t_i EI_i ) -?_j¦(1/t_j EE_j ) )
where
BW – body weight;
BW¬0 – initial body weight;
? – time scale of body weight gain/loss
t– time
EIi– Energy Intake of food of type

The food types are divided into different subgroups, e.g. at a larger scale, they include the known divisions - Fat, Carbohydrates, Proteins; but they can be further subdivided based on sources, e.g carbohydrates from sugar, starch, vegetables, dairy, and the like. In this equation, different types of energy intake (food intake) are summed up with a specific weighting (1/?i) that distinguishes each food item based on the above mentioned classification. A mechanistic interpretation of this weighting is the timescale of the metabolism of the food item wherein it is digested and contributes to the body weight. Including this weighting factor is a fundamental mechanism which distinguishes carbohydrates from say cane sugar as against those obtained from eating fresh fruits. Another interpretation of this is the allowance of differences in the metabolism of different type of food groups depending upon its sources.

Finally, this weighting is also specific to each user depending upon his/her own metabolism (which depends again on factors like age, weight, sex, etc). Initially, theseweights are established using averaged benchmarked data for users similar to the individual’s demographic attributes. As more specific individual data on users weight and food intake is gathered, we use a standard non-linear regression algorithm to establish and estimate the user specific timescales for each of the food groups, and update the parameters. This model then can be used to get newer predictions on the user’s weight loss journey.

The term EEj – in the equations represents energy expenditure (physical activity). Energy expenditure is also a sum of different activities (just like the energy intake), which user performs, both actively as well as passively. A workout that includes running, cycling, yoga, or the like is an active activity, while routine activities (breathing, working, sitting, sleeping) are considered passive.

Like in the energy intake, ?j?–refers to the timescale of contribution of the activity to the user’s body weight. It is further customized to reflect the individual user’s specific time constant of weight loss/gain. The total term is a collection of all user activities, as per his or her logs in the mobile app and/or inference from connected wearable devices. Each activity can affect the body weight differently and hence is assigned a specific time constant, which is further inferred and estimated for each individual based on collected data. Initially, in absence of the data, an average measure from data of similar users is used.

Given the measurement of the body weight following the prescribed change in food intake, the system and method can estimate the expected objective achievement as per the calculation above and assuming a particular values of the parameters. Comparing that with the measured body weight from user’s logs, the system and method can then make an updated estimate of parameters so that the expected body weight matches with the observed body weight. That estimated parameters can be now used to make a refined metabolic profile of the user which can then be further used to get better calorific targets. Every time a new weight measurement is available; the system and method can make estimate the parameters again. The final value of parameters used for targets is averaged across all measurements, thereby not giving any undue pivot to a particular data point.

A second correlation engine (CE2), governed by a second rule engine, correlates a second synthesised profile of a user with a satiety profile of a food item to determine a score of relevancy of satiety for a user with respect to the food item. This satiety score is a component of a satiety quotient whilst recommending a food item.

For relevance, an estimation of likelihood of a potential recommendation of a food item from the first storage device is correlated with respect to user taste profile.

Once the user taste profile is created, the system and method can compute the likelihood of any food item to be in accordance with a user’s taste preference by computing the distance of its co-ordinates in the state space to any of the items liked by the user in the corresponding taste profile. Mathematically, it is accomplished by computing the Euclidian distance of the new food item from each of the co-ordinates of the users taste profile, and then using the minimum of that. According to this system and method, it imposes the following condition to estimate the probability of any food item i to be within the users taste preference:
1-p_i=min?{d(t_j,t_i)}where
pi is the probability of user liking the food item i
t = taste vector for items-
j = items in the user taste preference coming from historical data and user logs

The system and method is further configured to incorporate the food items texture and odour to characterize the user’s preferences. Like with the taste space described above, both odour and texture will have their own subspace on which each of the food item would be profiled and stored. The user preferences and history will also be stored accordingly.

In accordance with an embodiment of this invention, a health parameter configuration engine (HE) is configured to create a health profile for each food item. A health vector mapping engine is configured to map said health vectors for a health profile for a food item in order to provide relevant recommendations from said recommendation engine.Each food item is defined by health vectors and stored in a health storage device, each of the health vectors being correlated to a food item. A plurality of health vectors for a health profile for a food item. These health vectors and correlations are further used in order to map it to recommendations provided to a user.

A third correlation engine (CE3), governed by a third rule engine, correlates a third synthesised profile of a user with a health profile of a food item to determine a score of relevancy of health for a user with respect to the food item. This health score is a component of a health quotient whilst recommending a food item.

In accordance with an embodiment of this invention, a recommendation engine (RE) configured to provide food item output from the first storage device based on rules configured by a rule engine. The rule engine receives inputs from first storage device, second storage device, third storage device, first selection mechanism, second selection mechanism, input mechanism, metabolic profiling mechanism, calorific computation engine, meal evaluation engine, activity monitoring module, a taste parameter configuration engine, first correlation engine, satiety parameter configuration engine, second correlation engine, health parameter configuration engine, health parameter configuration engine, and third correlation engine; in order to output a food item from the first storage device, which recommendation is pertinent to a user in terms of satiety quotient, taste quotient, and health quotient.The recommendation engineis a function of a first correlation engine, a second correlation engine, and a third correlation engine. In other words, therecommendation engine outputs a signal which comprises at least a content item having a cumulative strength of a signal corresponding to a health quotient, a signal corresponding to a taste quotient, and a signal corresponding to a satiety quotient.

The core of this invention is the automatic recommendation of the user meals that account for the user taste and metabolic profile. This is accomplished by a two stage mechanism. At the first step, the mechanismcomputes the typical food items that can be combined for a meal. The choice of food items is based on a mathematical scheme that weighs the taste and the nutritional aspects along with the user preferences. At the second step, the meal combination is scaled to meet the calorific targets as per the estimations of the individual’s metabolic profile.

Further details of the two steps are given below:

The typical servings of food groups that should be present in a diet are outlined by nutritional science. This system and method is adapted and used as a criteria to combine food items so that the combination should represent each food group adequately (grains, fruits and vegetables, proteins, fats and dairy, etc). Once this is accomplished, it narrows down the items in the combinations based on the user taste profile and other preferences. These final combinations then form the user meal recommendations which are stored in a user specific storage device that is used to pick and recommend meals for the user.

Figure 4 illustrates a flowchart depicting steps of a providing a content item relating to a food item recommendation.

A user’s metablic profile (406) is also input along with target (408) to be achieved.
A user inputs data relating to a food item (402) which is considered. Using this input, the food item paired (408) with given food item using pairing model (410) as explained in Figure 3 and a stored user profile (412) is used in order to generate (414) an ordered priority of list of items paired with the selected food item. From present items (416), a combination is generated (418) which allows the system to generate nutrient reachability (422) for a particular meal or a set of meals. Data relating to content items corresponding to food items is listed (424) in order of decreasing importance for a particular user in correlation with the user’s user profile (412) and the system’s pairing model (410). Each set of recommendations, generated at this stage, is classified (426) in terms of lower and upper bounds with respect to a user’s profile and with respect to a user’s target. This is then used in scaling (428) of content data corresponding to scaling of food items which is further used in the various scoring engines (430). Simultaneously, content items relating to targets to scale combination (420) is also used in scaling (428) of content data corresponding to scaling of food items which is further used in the various scoring engines (430). Each score or a cumulative score is checked (432) against threshold values and then output is generated.

In at least one embodiment, a gamification feature is provided that can track user’s good food habits and allows the user to redeem it for popular healthy activities.

Another aspect of the present invention allows a user to optimize meals.

This feature provides the user with real time optimization of the meal components based on his/her taste and metabolic profile. The user enters current meal components and uses the system of this invention to evaluate it. The system looks for components of the meals in terms of elementary nutritional measures (carbohydrates, proteins, vitamins, minerals, fibers, sugar etc) and then checks that again the daily targets set based on the user’s metabolic profile (and/ or meal plans). If the target is not met, the system makes modifications to the meal items by adding new items/scaling the portions to make the meal achieve the user-specific meal targets.

Yet another aspect of the present invention allows a user to build meals.

This feature provides the user with recommendations based on specific user-provided ingredient list and also provide with possible recipes. The recommendations are based on the user’s taste and metabolic profiles, as before.

The TECHNICAL ADVANCEMENT of this invention lies in providing a system and method which provided personalised recommendations relating to food items. The invention alleviates the difficulty of adhering to strict regimes, and also provides personalised recommendations and scores by incorporating a user’s tastes and preferences while adhering to dietary targets that are supplied by the user and / or computed directly based on available knowledge. Further, the invention provides:
System, Method, and its implementation on a mobile/web device to compute the user taste profiles and estimate the probability of ‘likeability’ of a new food item based on his previous history
System, Method, and its implementation on a mobile/web device to learn user’s metabolic profile that can adjust the individual diet targets in a dynamic manner
System, Method, and its implementation on a mobile/web device to integrate the user diet targets with the user taste profiles/preferences and physical activity and provide diet recommendations from food storage device that meet the dietary goals. Alternatively, it will also provide the rank order of the user meals, when queried to provide the user with a real time check on the suitability of the meal items, and also the appropriate portion size
System, Method, and its implementation on mobile/web device to allow the user to build and evaluate meals starting from specific food items/ingredients so as to maximize their health benefits for the user and also the probability of alignment of the meal taste and satiety to user’s preferences

In the advanced stage of the application, the system and method of this invention will incorporate the dynamics of blood glucose increase following the food intake and the subsequent lipogenesis to directly address the timing of the food intake and the generation of fat tissue. This model (in form of dynamic differential equations) will be personalized for the user based on the time course of weight changes after a given diet.

The most effective manifestation of this invention is a wearable device which comprises all these mechanisms.

While this detailed description has disclosed certain specific embodiments for illustrative purposes, various modifications will be apparent to those skilled in the art which do not constitute departures from the spirit and scope of the invention as defined in the following claims, and it is to be distinctly understood that the foregoing descriptive matter is to be interpreted merely as illustrative of the invention and not as a limitation.
,CLAIMS:WE CLAIM,

1. A system and method for providing personalised dietary recommendations based on multiple parameters, aid system comprising:
- a first storage device networked within said system, said first storage device configured to store content items pertaining to a food item;
- a second storage device networked within said system, said second storage device configured to store content items pertaining to calorific value of food items of said first storage device;
- a third storage device networked within said system, said third storage device configured to store content items pertaining to geographic location of food items of the first storage device;
- an attribute manager configured to determine and store attribute-related content items pertaining to food items stored in said first storage device;
- a first selection mechanism configured to prompt a user to select at least a choice of food item pertinent to said user;
- a second selection mechanism configured to prompt a user to select at least a food item that a user has ingested along with time, date, and serving size;
- an input mechanism configured to prompt a user to input said user’s details relating to at least one of height data, weight data, age data, gender data, location data, ethnicity data, genomic data, genetic data, and pertinent body data, characterised in that, said input mechanism comprising an ecosystem of wearables configured to provide distributed nodes as input mechanisms;
- a metabolic profiling mechanism is configured to read and store a metabolic profile of a user;
- a first dynamic GUI generator configured to generate a user-specific dynamic GUI corresponding a user’s first synthesized profile correlative to said user’s taste quotient, said first dynamic GUI generator being communicably coupled to said first selection mechanism, said second selection mechanism, said input mechanism, and said metabolic profiling mechanism;
- a second dynamic GUI generator configured to generate a user-specific dynamic GUI corresponding a user’s second synthesized profile correlative to said user’s satiety quotient, said second dynamic GUI generator being communicably coupled to said first selection mechanism, said second selection mechanism, said input mechanism, and said metabolic profiling mechanism;
- a third dynamic GUI generator configured to generate a user-specific dynamic GUI corresponding a user’s third synthesized profile correlative to said user’s health quotient, said third dynamic GUI generator being communicably coupled to said first selection mechanism, said second selection mechanism, said input mechanism, and said metabolic profiling mechanism;
- a taste parameter configuration engine configured to create a taste profile for each food item, characterised in that, each food item being defined by taste vectors and stored in a taste-related storage device, each of said taste vectors being correlated to a food item;
- a first correlation engine, governed by a first rule engine, correlating a first synthesised profile of said user with a taste profile of a food item to determine a score of relevancy of taste for said user with respect to said food item;
- a satiety parameter configuration engine configured to create a satiety profile for each food item, characterised in that, each food item being defined by satiety vectors and stored in a satiety-related storage device, each of said satiety vectors being correlated to a food item;
- a second correlation engine, governed by a second rule engine, correlating a second synthesised profile of said user with a satiety profile of a food item to determine a score of relevancy of satiety for said user with respect to said food item;
- a health parameter configuration engine configured to create a health profile for each food item, characterised in that, each food item being defined by health vectors and stored in a health-related storage device, each of said health vectors being correlated to a food item;
- a third correlation engine, governed by a third rule engine, correlating a third synthesised profile of said user with a health profile of a food item to determine a score of relevancy of health for said user with respect to said food item; and
- a recommendation engine configured to provide a food item output from the first storage device based on outputs from said first correlation engine said second correlation engine, and said third correlation engine.

2. A system as claimed in claim 1 wherein, said system comprising a calorific computation engine configured to receive a user’s weight goal in order to determine at least an appropriate cut in the calorific intake of the user and provide said data to said first correlation engine, said second correlation engine, and said third correlation engine in order to compute a staggered time-defined, goal-defined calorific data per user per time period per goal.

3. A system as claimed in claim 1 wherein, said system comprising a meal evaluation engine configured to receive data from said second selection mechanism in order to output data relating to a user’s ingested meal into pre-defined attribute content items.

4. A system as claimed in claim 1 wherein, said system comprising an activity monitoring module from said ecosystem of wearables configured to monitor physical activity of a user and store it in terms of activity data items.

5. A system as claimed in claim 1 wherein, said system comprising a taste vector mapping engine configured to map said taste vectors for a taste profile for a food item in order to provide relevant recommendations from said recommendation engine.

6. A system as claimed in claim 1 wherein, said system comprising a taste vector mapping engine configured to map said taste vectors for a taste profile for a food item in order to provide relevant recommendations from said recommendation engine, characterized in that, said taste vector mapping engine being configured to map each food item in a six-dimensional space array, each array correlative to at least a taste type and each item of each array of each taste type being populated with an intensity signal for that taste type.

7. A system as claimed in claim 1 wherein, said system comprising a taste vector mapping engine configured to map said taste vectors for a taste profile for a food item in order to provide relevant recommendations from said recommendation engine, characterized in that, said taste vector mapping engine being configured to map each food item in a six-dimensional space array, each array correlative to at least a taste type and each item of each array of each taste type being populated with an intensity signal for that taste type, further characterized in that, said taste vectors being weighted and normalised taste vectors based on ingredients per food item.

8. A system as claimed in claim 1 wherein, said system comprising a satiety vector mapping engine configured to map said satiety vectors for a satiety profile for a food item in order to provide relevant recommendations from said recommendation engine.

9. A system as claimed in claim 1 wherein, said system comprising a satiety vector mapping engine configured to map said satiety vectors for a satiety profile for a food item in order to provide relevant recommendations from said recommendation engine, characterized in that, said space vector mapping engine being configured to map each food item with a user-specific calorific target based on data from a group consisting of food ingredient composition and time of ingestion.

10. A system as claimed in claim 1 wherein, said system comprising a satiety vector mapping engine configured to map said satiety vectors for a satiety profile for a food item in order to provide relevant recommendations from said recommendation engine, characterized in that, said space vector mapping engine being configured to map each food item with a user-specific calorific target based on data from a group consisting of food ingredient composition and time of ingestion, further characterized in that, said satiety quotient being correlative to a body weight and body type.

11. A system as claimed in claim 1 wherein, said system comprising a health vector mapping engine configured to map said health vectors for a health profile for a food item in order to provide relevant recommendations from said recommendation engine.

12. A system as claimed in claim 1 wherein, said first storage device being a set of relationally-defined interconnected devices¸ wherein the set of devices cumulatively comprises at least a content item pertaining to an identity of a food item per se, at least a content item relating to recipe of the food items, at least a content item relating to ingredients of the food item, at least a content item relation to at least a nutrient of the food item, at least a content item relating to a pre-defined parameter of the food item.

13. A system as claimed in claim 1 wherein, said second storage device being a set of relationally-defined interconnected devices¸ wherein the set of devices cumulatively comprises at least a content item pertaining to calorific values of food items as well as ingredients of the food item.

14. A system as claimed in claim 1 wherein, said third storage device being a set of relationally-defined interconnected devices¸ wherein the set of devices cumulatively comprises at least a content item pertaining to geographic location(s) of the food items, ingredients of the food item, cultural attributes of the food items.

15. A system as claimed in claim 1 wherein, said input mechanism being configured to receive user inputs, wherein each input being correlated to a signal comprising a content item fetched from a group of storage devices comprising said first storage device, said second storage device, said third storage device, along with a signal comprising data from said attribute manager, and a signal comprising data correlating to a time parameter.

16. A system as claimed in claim 1 wherein, said recommendation engine is governed by a rule engine, said rule engine receiving inputs from said first storage device, said second storage device, said third storage device, said first selection mechanism, said second selection mechanism, said input mechanism, said metabolic profiling mechanism, said calorific computation engine, said meal evaluation engine, said activity monitoring module, said taste parameter configuration engine, said first correlation engine, said satiety parameter configuration engine, said second correlation engine, said health parameter configuration engine, said third correlation engine, in order to output a signal signal which comprises at least a content item correlating to a food item, said output signal having a cumulative strength of a signal corresponding to a health quotient, a signal corresponding to a taste quotient, and a signal corresponding to a satiety quotient.

Dated this 26th day of September, 2017.

CHIRAG TANNA
of INK IDEE
APPLICANT’s PATENT AGENT