FORM 2
THE PATENT ACT 1970 &
The Patents Rules, 2003 COMPLETE SPECIFICATION
(See section 10 and rule 13)
1. TITLE OF THE INVENTION:
DYNAMIC PRODUCT TAXONOMY AND HEURISTIC APPROACH BASED RECOMMENDATION ENGINE FOR MOBILE COMMERCE
Name Nationality Address

Applicant:

GUJARAT UNIVERSITY AN INDIAN NATIONAL NAVRANGPURA, AHMEDABAD, GUJARAT-380009, INDIA.

Inventers:

DR. SHANTI VERMA AN INDIAN NATIONAL L.J. INSTITUTE OF COMPUTER APPLICATIONS, GUJARAT TECHNOLOGICAL UNIVERSITY, AHMEDABAD-382210, GUJARAT, INDIA. E-Mail: verma.shanti@gmail.com

DR. KALYANI PATEL (ASSISTANT PROFESSOR) AN INDIAN NATIONAL K.S. SCHOOL OF BUSINESS MANNAGEMENT, GUJARAT UNIVERSITY, AHMEDABAD-380009, GUJARAT, INDIA. E-Mail :patelkalyani05@gmail.com

REAMBLE TO THE DESCRIPTION
COMPLETE
The following specification Invention. Particularly describes the invention and the manner in which it is to be performed.


FIELD OF THE INVENTION
The invention “DYNAMIC PRODUCT TAXONOMY AND HEURISTIC APPROACH BASED RECOMMENDATION ENGINE FOR MOBILE COMMERCE “is relate to techniques for mapping products between different taxonomies, information filtering and recommendation systems.
BACKGROUND OF THE INVENTION
Mobile Commerce or online shopping using handheld devices are now in fashion and used by every age people. Mobile commerce provides users to shop anything from anywhere. Recommendation systems are the important part of mobile-commerce which helps users to shop the next item with respect to the item purchased before with ease. There are various types of recommendation systems; Collaborative filtering based is one of them used by us in this research work. Recommendation systems are used by every online shopping application to increase the revenue of business and provide customer satisfaction. Study of recommendation engine for mobile-commerce site using heuristic approach has been undertaken by keeping us the following motivational points:
1. In India, younger age people (below 35) are 75% which plays important role in the growth recommendation system of mobile-commerce.
2. Accessibility, entertainment, reliability, mobility, externality and reciprocity plays important role in success of Mobile commerce.
3. By the end of year 2020 it will be expected that 10 billion Smartphone’s and 1 billion tablet users worldwide [16].
4. Traditional Retail can serve only most popular products. Online can serve much more products, but it’s overwhelming for customers.
5. Main goals of recommendation system are cross-sale and save customer time.
This research is to provide fast and precise product recommendation to mobile commerce users. The performance of recommendation engine is based on the

product taxonomy. In this research we have used weighted product taxonomy which provides fast search of product. In our research we mainly focus on customer behavioral and navigational factors to calculate weight of product taxonomy and provide recommendation. Our main contributions to in this research are outlined as follows:
Use weighted product taxonomy and apply heuristic search algorithm to provide fast search. In calculation of neighborhood, we have used Nearest Neighborhood algorithm with person correlation coefficient. This would helpful to provide better cluster formation for precise product recommendation. The thesis is structured in seven chapters.
The research preamble which includes introduction, aim, motivation and benefits of research. It also includes the steps carried out with description to achieve objectives of research. The literature review of topics covered in this thesis. These topics are recommender system on the basis of scope and usage, Collaborative filtering techniques, role of ontology in recommendation system, and various machine learning algorithms used in recommendation engine. This chapter also provides the work done by other researcher in the field of recommendation engine. Chapter 3 elaborates the primary survey conducted by us to know the Indian customer behavior while doing online shopping. The main objective to conduct primary survey is to know the most popular mobile- commerce application in India.
We use the same application dataset for analysis. The research methodology used by us in this research. It includes the description of various steps to recommend product to customer by proposed technique. This chapter also includes the pseudo code and flow chart of proposed heuristic search algorithm. Chapter 5 provides all the steps which are necessary for data analysis. It includes hypothesis formation, finding appropriate statistical test, data collection, data generation for analysis using proposed heuristic algorithm, and its interpretation of result. Chapter 6 elaborates results and discussion on the basis of precision, recall and accuracy values and precision-recall curve. The contribution in research in steps followed by

work flow diagram. Future work, long term and short term goals are also discussed in this chapter.
The contain the questionnaire of primary survey which is useful to find the most popular mobile commerce application in India. Appendix II contains the raw data collected from primary survey which is helpful to know the Indian consumer behavior while doing online shopping. The product category tree structure for genre of “Clothing” and “Watches” which are used for generation of weighted product taxonomy. The sample data set from consumer behavioral patterns collected from kaggle.com. One technique commonly used by recommendation services is known as content-based filtering. Pure content-based systems operate by attempting to identify items which, based on an analysis of item content, are similar to items that are known to be of interest to the user. For example, a content-based Web site recommendation service may operate by parsing the user's favorite Web pages to generate a profile of commonly-occurring terms, and then use this profile to search for other Web pages that include some or all of these terms.
Content-based systems have several significant limitations. For example, content-based methods generally do not provide any mechanism for evaluating the quality or popularity of an item. In addition, content-based methods generally require that the items include some form of content that is amenable to feature extraction algorithms; as a result, content-based systems tend to be poorly suited for recommending movies, music titles, authors, restaurants, and other types of items that have little or no useful, parsable content.
Existing source content selection systems are quite ineffective in supporting content searches much beyond using artist, collection, and title. Users therefore typically confine their searches to just those media items that are independently known to them or are aware of through other sources of media information. These other sources are typically sufficient to provide indications of whether and which segments of the general population might appreciate particular content items. No

indication is given and none can be reliably inferred as to whether a particular user will enjoy or appreciate a given item.
There is, at least for entertainment media content, some acceptance of the belief that a user's appreciation of particular content items can suggest the user's likely appreciation of other content titles. Systems built to exploit this belief have met with limited results. One known system, apparently a neural-net based expert system, determines and provides recommendation of other content titles based purely on the similarities between users without considering the relationships between the music items from a content or contextual point of view. These systems have the disadvantage that they require an initial "teaching" period where the recommendations given to users are likely to be inaccurate. Another disadvantage is that the user does not understand the reasoning behind the recommendations and therefore does not trust the recommendations. The absence of confidence in whatever recommendations are given directly reduces the utility of the system. Additionally, such systems tend to generate recommendations that reflect the lowest common denominator between broad users tastes.
As a result, these systems typically provide recommendations reflecting potential appreciation within a single generic style, such as only 1 980's pop music. These systems do not appear to be effectively capable of providing recommendations across a diverse range of music, such as Death Metal and Classical. Another known system recommends particular content items based on the given content or style of the item. Such systems are generally established by - 3 - hand, requiring a broad, yet detailed, understanding of each media item.
Therefore, a general purpose of the present invention to provide a system that combines content-based filtering and progressively refined collaborative- based filtering to deliver a set of media item recommendations that are consistent with a user's personal media content interests. This purpose is achieved in the present invention by providing a system and method of providing media content recommendations through a computer server system connected to a network

communications system. The computer server system preferably has access to a first database of media content items including media content and related information and a media content filter identifying and providing qualifying attribute relationship data for media content items within the first database.
The media content recommendations are particularly tailored to the personalized interests of a user through sequence of steps including presenting media content items through a network-connected interface to the user for review and consideration of potential personal interest, monitoring the consideration of the media content items implied through the user directed navigation among the presented media content items and user requests for related information; collecting the monitored data to develop a user weighted data set reflective of the user's relative consideration of the media content items; and evaluating the user weighted data set in combination with the media content filter to identify a set of media content items accessible from the first database for re- presentation to the user. Thus, the operation of the present system reflects the consideration that media content items, such as music, video, and other forms of content, can be interrelated based on multiple characterizing attributes.
The strength of these characterizing attributes, or similarities, is used to further define these content- based relationships, even as between quite different forms or types of media content. An additional aspect of the operation of the present invention allows for the progressive or continuing collaborative, including self-collaborative, development of such content-based relationships. An advantage of the present invention, therefore, is that the provided combination of content and collaborative recommendation systems enables the delivery of recommendations that are particularly tailored to the personalized interests of a user. Another advantage of the present invention is that the system flexibly determines a scope of applicable similarities between a particular and other users and recommends items within the applicable scope.

The invention is that the self-collaborative relationships developed for individual users of the system permit the development of individualized recommendations even where the group collaborative relationships reflect the choices of users with highly diverse media content interests. Still another advantage of the present invention is that the system enables multi-level media content relationship information to be captured and used as data evaluate able in providing particularized media content item recommendations. Yet another advantage of the present invention is that implicit and explicit collaborative data is captured from and in consideration of particular users, supporting both the continuing development of both group and personal interest profiles. The implicit collaborative data is advantageously obtained from a user's self-directed actions of reviewing and considering different media content items.
Thus, the selection of items to review and the length and nature of the consideration of such items inferentially reflects the user's relative interest in particular media content items. Confidence levels in the inferences drawn can also be developed and refined through the continued monitoring of user actions in reviewing and considering the same and closely similar media content items. The explicit information provided by users regarding the level and nature of their interest in different media content items provides high-confidence information that can be incorporated into the group and individualized collaborative data.
Conventional retailer websites allow shoppers to browse through a wide variety of products available for sale online. Each retailer website typically hosts multiple product listing webpages that offer various products for sale. Moreover, each retailer website generally maintains its own inventory of products. These different inventories may be stored using different taxonomies for each retailer website. It can be difficult, therefore, to compare products between websites or perform other comparative functions because it can be difficult to determine with precision whether a particular product on one retailer website is identical to a particular product on another retailer website.

PRIOR ART STATEMENT
US8078604B2 * 2007-03-192011-12-13 Microsoft Corporation
Identifying executable scenarios in response to search queries.
US7818341B2 *2007-03-192010-10-19 Microsoft Corporation
Using scenario-related information to customize user experiences.
US20090219574A1 *2007-03-192009-09-03 Dnp Photo Imaging
America Corporation System and method for the preparation of identification cards utilizing a self-service identification card station.
US8510453B2 *2007-03-212013-08-13 Samsung Electronics Co.,
Ltd. Framework for correlating content on a local network with information on an external network.
US8065675B2 2007-03-212011-11-22 Yahoo! Inc.In-page
Installer.
US8799250B1 *2007-03-262014-08-05 Amazon Technologies,
Inc. Enhanced search with user suggested search information.
US8560398B1 2007-03-292013-10-15 Amazon Technologies,
Inc. Method and system for providing item recommendations.
US7743059B2 *2007-03-302010-06-22 Amazon Technologies,
Inc. Cluster-based management of collections of items.
US8095521B2 *2007-03-302012-01-10 Amazon. Technologies,
Inc. Recommendation system with cluster-based filtering of recommendations.
US7881984B2 *2007-03-302011-02-01 Amazon Technologies,

Inc. Service for providing item recommendations.
US7966225B2 *2007-03-302011-06-21 Amazon Technologies Inc.
Method, system, and medium for cluster-based categorization and presentation of item recommendations.
US8117193B2 2007-12-212012-02-14 Lemi Technology,
LlcTunersphere.
US8316015B2 2007-12-212012-11-20 Lemi Technology,
LlcTunersphere.
US8060525B2 *2007-12-212011-11-15 Napo Enterprises,
LlcMethod and system for generating media recommendations in a distributed environment based on tagging play history information with location information.
US7904530B2 *2008-01-292011-03-08 Palo Alto Research
Center Incorporated Method and apparatus for automatically incorporating hypothetical context information into recommendation queries.
US9704202B1 2008-02-182017-07-11 Sprint Communications
Company L.P. Blending device-based recommendations.
US8554891B2 *2008-03-202013-10-08 Sony Corporation
Method and apparatus for providing feedback regarding digital content within a social network.
US8725740B2 2008-03-242014-05-13 Napo Enterprises,
Locative playlist having dynamic media item groups.
US20090254633A1 *2008-04-032009-10-08 Olive Bentley

Methods, systems, and computer program products for distributing profile-based advertisement content and user identification-tagged media content.
US20090259621A1 *2008-04-112009-10-15 Concert Technology
Corporation Providing expected desirability information prior to sending a recommendation.
US8484311B2 2008-04-172013-07-09 Eloy Technology,
LlcPruning an aggregate media collection.
US8204883B1 2008-04-172012-06-19 Amazon
Technologies, Inc. Systems and methods of determining genre information.
US20090271246A1 *2008-04-282009-10-29 American Express Travel
Related Services Company, Inc.Merchant recom.
US10217147B220 14-09-122019-02-26E bay Inc. Mapping
products between different taxonomies.
US10319019B2 2016-09-142019-06-11 EBay Inc. Method,
medium, and system for detecting cross-lingual comparable listings for machine translation using image similarity Family To Family Citations.
US20180101893A1 *2016-10-112018-04-12 EBay Inc. Systems
and methods to select a product title.
US20180285957A1 *2017-03-312018-10-04 Ten-X, LlcRecomme
This application is a continuation of U.S. patent application Ser. No. 09/850,263, filed May 7, 2001 now U.S. Pat. No. 7,113,917, which is a continuation of U.S. patent application Ser. No. 09/156,237, filed Sep. 18, 1998 (now U.S. Pat. No. 6,317,722).

This international application claims the benefit of priority from U.S. Provisional Application Serial No. 62/049,590, entitled "MAPPING.
PRODUCTS BETWEEN DIFFERENT TAXONOMIES," filed on September 12, 2014, and U.S. Serial No 14/517,505, filed on October 17, 2014, which are hereby incorporated by reference herein in their entirety.
OBJECTIVES OF THE INVENTION
The objective of the invention is to wherein the data repository comprises a mapping structure that associates related items. The first component uses the data values to select item pairs to include in the mapping structure. The other objective of the invention is to wherein the first component is configured to exclude, from the data repository, data values that do not satisfy a selected condition. The other objective of the invention is to wherein the input list includes at least one purchased item that was purchased by the user on a purchase date, and the second component is configured to take the purchase date into consideration in determining how much weight to give to the purchased item in generating the output list of items.
The other objective of the invention is to wherein the input list includes at least one item to which the user has explicitly assigned a rating, and the second component is configured to take the rating into consideration in generating the output list of items.
The other objective of the invention is to wherein the first component is configured to additionally take item viewing histories of users into consideration in generating the data values. The second component generates the input list of items, at least in part, by identifying items viewed by the user during browsing of an electronic catalog.
The other objective of the invention is to wherein incorporating the item recommendations into a page comprises incorporating the item recommendations into a shopping cart page that displays contents of an electronic shopping cart of the user.

The other objective of the invention is to wherein the item recommendations
are based on said contents of the electronic shopping cart.
The other objective of the invention is to wherein the method is performed
in its entirety by a web server.
The other objective of the invention is to wherein the recommendation
service is external to the web server, and the method comprises the web
server communicating with the recommendation service to obtain the item
recommendations. . The other objective of the invention is to wherein the method is performed
in its entirety by a web server system that is separate from the
recommendation service. . The other objective of the invention is to further comprising generating the
item recommendations with the recommendation service. . The other objective of the invention is to wherein the item recommendations
are personalized for the user based at least partly on an item ratings profile
of the user. . The other objective of the invention is to wherein the item recommendations
are personalized for the user based at least partly on past item purchases of
the user. . The other objective of the invention is to wherein the item recommendations
are personalized for the user based at least partly on recorded item viewing
activities of the user. . The other objective of the invention is to wherein said profiling system
provides profile data combinable with said weighted content relationship
information relative to a characteristic attribute of the content selectable by
said user of said client system. . The other objective of the invention is to wherein said profiling information
reflects explicit indications of interest and implicit indications of interest in
particular instances of content. . The other objective of the invention is to wherein said implicit indications of
interest are derived from the selection and duration of sampling of
particular instances of content.

The other objective of the invention is to wherein the at least one feature
related to text comprises a title.
The other objective of the invention is to wherein the at least one feature
related to text comprises a description.
The other objective of the invention is to wherein the at least one feature
related to the item listing, but other than a feature related to an image and a
feature related to text, comprises a Universal Product Code (UPC).
The other objective of the invention is to wherein the creating a mapping
includes using a machine learning model to create the mapping.
The other objective of the invention is to wherein the machine learning
model is based on a random forest model.
The other objective of the invention is to wherein the machine learning
model is based on a logistic regression model.
The other objective of the invention is to further comprising: receiving
feedback from a user with regard to one or more of the item listings; and
using the feedback in the machine learning model to update the machine
learning model; and creating a new mapping based on the update to the
machine learning model.
The other objective of the invention is to including, for each of the obtained
item listings, extracting at least one feature from an image associated with
the item listing, wherein the creation of the mapping between item listings
in the first taxonomy and item listings in the second taxonomy is also based
on the at least one extracted feature from the images associated with the
item listings. A machine-readable medium carrying instructions, which
when implemented by one or more machines, cause the one or more
machines to perform operations comprising: obtaining item listings from a
first database storing a first taxonomy and item listings from a second
database storing a second taxonomy; for each of the obtained item listings,
extracting a plurality of features, including at least one feature related to an
image associated with the item listing and at least one feature related to text
associated with the item listing; and creating a mapping between item
listings in the first taxonomy and item listings in the second taxonomy based

on the plurality of features extracted by the feature extraction component,
wherein the mapping identifies which item listings in the first taxonomy
correlate to a same product as which item listings in the second taxonomy.
The other objective of the invention is to further comprising performing
computing at least one additional feature from the plurality of features. A
machine readable medium carrying machine readable instructions for
controlling a machine to carry out the method.
The other objective of the invention is to further comprising adjusting the
number of scores combined in the combined score to reduce the impact of
outliers on the normalized scores.
The other objective of the invention is to wherein combining the scores for
at least some of the candidate recommendations comprises summing the
scores.
The other objective of the invention is to further comprising assigning
weights to candidate recommendations received from the first and second
recommenders.
The other objective of the invention is to wherein calculating normalized
scores based at least in part on the combined score and the scores for at least
some of the candidate recommendations comprises dividing each candidate
recommendation score by the combined score.
The other objective of the invention is to wherein the moving average
comprises an exponential moving average.
SUMMARY OF THE INVENTION
items available in mobile-commerce site is arranged in a hierarchical manner ategory of items. The hierarchical structure of product is called product
my. There are millions of products in product taxonomy so searching of ular product is very tedious task. Traditional recommendation systems use search algorithms like Breadth First Search (BFS) to search product in
my which is very slow as each item is visited. In the research work we uced a concept of weighted product taxonomy with heuristic search approach.

Weighted product taxonomy is a hierarchical structure of products as nodes joined with weight. Weights between nodes are calculated with the help of customer navigational information. In the research work we proposed a heuristic algorithm to search product in weighted product taxonomy. Heuristic is a greek word which meaning is guidance. So our proposed algorithm has one function which provides guidance to search the next node. The main objective of proposed heuristic algorithm is to provide optimal path with optimal time while searching product in taxonomy. We decided a null hypothesis that proposed heuristic algorithm traversal time is less than traditional BFS algorithm.
To find the most used mobile commerce application in India, we conduct a survey. The outcome derived from survey on the basis of 339 records, Flipkart is the most used application in India. We choose two product categories “Clothing” and “Watches” as a sample from Flipkart dataset. From different start and goal nodes of product taxonomy of Clothing and Watches, we note the traversal time in seconds. The traversal time is noted for proposed heuristic and BFS algorithm for both product categories. To test the null hypothesis, we use equal variance two sample independent t test with 95% level of significance. Since we have to check inequality one sided test value is used for interpretation.
The one sided t test value obtain from test is 0.24123 and 0.43008 for product category “Clothing” and “Watches” respectively. We have decided alpha is 5% or 0.05. According to statistical law ‘if calculated value is greater than alpha value’, Null hypothesis (H0) accepted. Results for product category “Clothing” shows that one tail t test value is 0.241234 which is greater than 0.05. Results for product category “Watches” shows that one tail t test value is 0.430089 which is also greater than 0.05. Both the results show the acceptance of null hypothesis. Interpretation can also be done on the basis of Critical t test value. Critical t value for 5% level of significance and 72 degree of freedom is 1.66. According to statistical law ‘if calculated value is less than critical t value’, Null hypothesis (H0) accepted. Results for product category “Clothing” shows that one tail t test value is0.241234 which is less than critical t value 1.66. Results for product category “Watches” shows that one tail t test value is 0.430089 which is also less than critical t value 1.66. These both

results show the acceptance of null hypothesis.
Results of ‘t’ test for product category Clothing and Watches shows the acceptance of null hypothesis so we can conclude that the mean traversal time of proposed (Heuristic) algorithm is less than the algorithm (BFS) used in conventional approach. It clearly shows that if we use weighted product taxonomy with heuristic search approach product recommendation is fast and it utilizes the system efficiency. We hope that in the digital world where number of online customers grows exponentially if mobile-commerce companies use our proposed approach they will be able to provide better and fast recommendation of products to their customers. This will help companies to increase their revenue and customer satisfaction.
An important benefit of the service is that the recommendations are generated without the need for the user, or any other users, to rate items. Another important benefit is that the recommended items are identified using a previously-generated table or other mapping structure which maps individual items to lists of “similar” items. The item similarities reflected by the table are based at least upon correlations between the interests of users in particular items. The types of items that can be recommended by the service include, without limitation, books, compact discs (“CDs”), videos, authors, artists, item categories, Web sites, and chat groups. The service may be implemented, for example, as part of a Web site, online services network, e-mail notification service, document filtering system, or other type of computer system that explicitly or implicitly recommends items to users. In a preferred embodiment described herein, the service is used to recommend works such as book titles and music titles to users of an online merchant's Web site.
In accordance with one aspect of the invention, the mappings of items to similar items (“item-to-item mappings”) are generated periodically, such as once per week, by an off-line process which identifies correlations between known interests of users in particular items. For example, in the embodiment described in detail below, the mappings are generating by periodically analyzing user purchase histories to identify correlations between purchases of items. The similarity between two items

is preferably measured by determining the number of users that have an interest in both items relative to the number of users that have an interest in either item (e.g., items A and B are highly similar because a relatively large portion of the users that bought one of the items also bought the other item). The item-to-item mappings could also incorporate other types of similarities, including content-based similarities extracted by analyzing item descriptions or content.
Recommendation Table:

Clothing Peppermint Blues Tops 0.02889 1 1 1 1
1 1 1 1 1 1 25 26 27 28
29 30 31 32 33 34 28 29 30 31
32 33 34 35 36 37 0.89286 0.86207 0.91892
Clothing Just In Time Fashions Dresses 0.023018


0.89656 0.89656 0.91892
Clothing Yepme Dresses 0.031398


0.9 0.93104 0.91892
Clothing Van Heusen Casual & Party Wear Shirts 0.034509


0.90323 0.96552 0.91892
Clothing Mykraft Dresses 0.034234


0.90625 1 0.91892
Clothing The Cranberry Club Dresses 0.027633


0.9091 1.03449 0.91892
Clothing Fashion Tree Formal Shirts 0.034092


0.91177 1.06897 0.91892
Clothing Peter England Formal Shirts 0.033102


0.91429 1.10345 0.91892
Clothing Gee & Bee Kurtis 0.022494


0.91667 1.13794 0.91892
Clothing Dumdaar.Com Tops 0.016945


0.91892 1.17242 0.91892

Product Category Tree Structure “Clothing”
Clothing >> Women's Clothing >> Lingerie, Sleep & Swimwear >> Shorts >> Alisha Shorts >> Alisha Solid Women's Cycling Shorts
Clothing >> Women's Clothing >> Sports & Gym Wear >> Swimsuits >> Carrel Swimsuits >> Carrel Printed Women's
Clothing >> Kids' Clothing >> Boys Wear >> Polos & T-Shirts >> dongli Polos & T-Shirts >> dongli Printed Boy's Round Neck T-Shirt (Pack of 4)
Clothing >> Women's Clothing >> Lingerie, Sleep & Swimwear >> Lingerie Sets >> Glus Lingerie Sets
Clothing >> Women's Clothing >> Fusion Wear >> Leggings & Jeggings >> Legging

Jegging >> FDT Legging Jegging
Clothing >> Men's Clothing >> Cargos, Shorts & 3/4ths >> Cargos >> Madcaps Cargos
Clothing >> Women's Clothing >> Ethnic Wear >> Fabric >> Lehenga Choli Material >> Indcrown Lehenga Choli Material >> Indcrown Net
Embroidered Semi-stitched Lehenga C...
Clothing >> Women's Clothing >> Formal Wear >> Waistcoats >> KAJCI Waistcoats >> KAJCI Embroidered Women's Waistcoat
Clothing >> Women's Clothing >> Formal Wear >> Waistcoats >> Pick Pocket Waistcoats >> Pick Pocket Embroidered Women's Waistcoat
Clothing >> Kids' Clothing >> Boys Wear >> Dungarees & Jumpsuits >> Dungarees >> Oye Dungarees >> Oye Boy's Dungaree
Clothing >> Men's Clothing >> Winter & Seasonal Wear >> Cardigans >> Roadster Cardigans >> Roadster Men's Zipper Solid Cardigan
Clothing >> Women's Clothing >> Western Wear >> Shirts, Tops & Tunics >> Shirts >> Mario Gotze Shirts >> Mario Gotze Women's Printed
Casual Orange Shirt
Clothing >> Men's Clothing >> Shirts >> Casual & Party Wear Shirts >> Discountgod Casual & Party Wear Shirts
Clothing >> Men's Clothing >> Shirts >> Casual & Party Wear Shirts >> Silver Streak Casual & Party Wear Shirts
Clothing >> Men's Clothing >> Jeans >> Reckler Jeans

Reality
Traversal time is Traversal time
less is
than 0.05 seconds greater than and
equal to 0.05 seconds

Proposed Heuristic
Algorithm provide fast
recommendation of product
than Conventional BFS
Type I error
algorithm for Mobile-

commerce site
Decision

Proposed Heuristic
Algorithm provide slow
recommendation of product
than Conventional BFS algorithm for Mobile-commerce site Type II error
For justification of proposed algorithm results, we have also apply precision-recall analysis with different threshold values for traversal time and discuss their results in this chapter. Precision and recall are the measures of accuracy. Precision deals with quality of results of algorithm while recall deals with quantity of results of algorithm. If the precision value is high of proposed algorithm we can conclude that algorithm gives more relevant results than irrelevant results. If the recall value is high, we can conclude that most results of the proposed algorithm are relevant. Precision and recall values are calculated on the basis of Type I and Type II error values. To understand the concept of Type I and Type II error let’s take an example. We have assumed the threshold value of traversal time is 0.05 seconds. If the null hypothesis decided by us is that the proposed algorithm provides fast recommendation as compare to conventional blind search algorithm with traversal time 0.05 seconds. This is only the assumption of us. What are the situations if assumption is not true?
Precision, recall and accuracy are calculated on the basis of formula based on True positive, True Negative, False Positive and False Negative True positive is the correct decision of us based on null hypothesis. True negative is the rejection of null

hypothesis. False positive is Type I error and false negative is the type II error. Precision is proportion of True Positive and actual results is sum of true positive and false positive. Recall is proportion of true positive and predicted results where a predicted result is sum of true positive and false positive. Accuracy is the proportion correct decisions and total outcomes. elaborates the precision, recall and accuracy formulas used by us.
Modules, Components and Logic:
Certain embodiments are described herein as including logic or a number of components, modules, or mechanisms. Modules may constitute either software modules (e.g., code embodied (1) on a non- transitory machine- readable medium or (2) in a transmission signal) or hardware-implemented modules. A hardware-implemented module is a tangible unit capable of performing certain operations and may be configured or arranged in a certain manner. In example embodiments, one or more computer systems (e.g., a standalone, client or server computer system) or one or more processors 602 may be configured by software (e.g., an application or application portion) as a hardware-implemented module that operates to perform certain operations as described herein.
In various embodiments, a hardware-implemented module may be implemented mechanically or electronically. For example, a hardware- implemented module may comprise dedicated circuitry or logic that is permanently configured (e.g., as a special-purpose processor, such as a field- programmable gate array (FPGA) or an application-specific integrated circuit (ASIC)) to perform certain operations. A hardware-implemented module may also comprise programmable logic or circuitry (e.g., as encompassed within a general-purpose processor or other programmable processor) that is temporarily configured by software to perform certain operations. It will be appreciated that the decision to implement a hardware-implemented module mechanically, in dedicated and permanently configured circuitry, or in temporarily configured circuitry (e.g., configured by software) may be driven by cost and time considerations.

Accordingly, the term "hardware-implemented module" should be understood to encompass a tangible entity, be that an entity that is physically constructed, permanently configured (e.g., hardwired) or temporarily or transitorily configured (e.g., programmed) to operate in a certain manner and/or to perform certain operations described herein. Considering embodiments in which hardware-implemented modules are temporarily configured (e.g., programmed), each of the hardware-implemented modules need not be configured or instantiated at any one instance in time. For example, where the hardware-implemented modules comprise a general-purpose processor configured using software, the general-purpose processor may be configured as respective different hardware-implemented modules at different times. Software may accordingly configure the processor 602, for example, to constitute a particular hardware-implemented module at one instance of time and to constitute a different hardware-implemented module at a different instance of time.
Hardware-implemented modules can provide information to, and receive information from, other hardware-implemented modules. Accordingly, the described hardware-implemented modules may be regarded as being communicatively coupled. Where multiple of such hardware-implemented modules exist contemporaneously, communications may be achieved through signal transmission (e.g., over appropriate circuits and buses that connect the hardware-implemented modules). In embodiments in which multiple hardware-implemented modules are configured or instantiated at different times, communications between such hardware-implemented modules may be achieved, for example, through the storage and retrieval of information in memory structures to which the multiple hardware-implemented modules have access. For example, one hardware-implemented module may perform an operation, and store the output of that operation in a memory device to which it is communicatively coupled. A further hardware-implemented module may then, at a later time, access the memory device to retrieve and process the stored output. Hardware-implemented modules may also initiate communications with input or output devices, and can operate on a resource (e.g., a collection of information).

The various operations of example methods described herein may be performed, at least partially, by one or more processors 602 that are temporarily configured (e.g., by software) or permanently configured to perform the relevant operations. Whether temporarily or permanently configured, such processors 602 may constitute processor-implemented modules that operate to perform one or more operations or functions. The modules referred to herein may, in some example embodiments, comprise processor-implemented modules.
Interpretation on the basis of Criticalt test value:
Criticalt value for 5% level of significance and 72 degree of freedom is 1.66. According to statistical law if calculated value is less than critical t value, Null hypothesis (H0) accepted.
Results for product category “Clothing” shows that one tail t test value is0.241234 which is less than critical t value 1.66. This shows the acceptance of null hypothesis.
Results for product category “Watches” shows that one tail t test value is 0.430089 which is less than critical t value 1.66. This shows the acceptance of null hypothesis.
Results of ‘t’ test for product category clothing and watches shows the acceptance of null hypothesis so we can conclude that the mean traversal time of proposed (Heuristic) algorithm is less than the algorithm (BFS) used in conventional approach. It clearly shows that if we use weighted product taxonomy with heuristic search approach product recommendation is fast and it utilizes the system efficiency.
The describes the steps used in data analysis. It includes 6 steps which are hypothesis formation, choose appropriate statistical test, decide alpha values, and gather sample dataset, statistical analysis and interpretation of results. This chapter also includes the summary for secondary data used for generation of

weighted product taxonomy on the basis of customer behavioral patterns. We conclude that proposed heuristic algorithm traversal time is less than conventional BFS algorithm traversal time for weighted product taxonomy.
Determine appropriate statistical test:
Statistical test refers to a function of sample observations. The computed value of statistical test determines the final decision regarding acceptance and rejection of null hypothesis (H0). If the value of the statistical test falls in the critical region1, the null hypothesis rejected. If the value of the statistical test does not fall in the critical region, the null hypothesis accepted.
There are various types of statistical test such as Z test, t test, Chi-square and F test. On the basis of sample size, nature of data and application of test, we decide which statistical test is appropriate. Z test is used when sample size is large or >30. t test is also known as Student 't' test used for small sample size <=30 where selection of samples is independent t test are of different types. It is very important to choose correct t test. If the correct test is not chosen by us, hypothesis interpretation may be wrong
BRIEF DESCRIPTION OF THE DIAGRAM
FIG. 1: illustrates a Web site which implements a recommendation service which
operates in accordance with the invention, and illustrates the flow of information
between components.
FIG. 2: illustrates a sequence of steps that are performed by the recommendation
process of FIG. 1 to generate personalized recommendations.
FIG. 3: illustrates a sequence of steps that are performed by the table generation
process of FIG. 1 to generate a similar items table, and illustrates temporary data
structures generated during the process.
FIG. 4: is a Venn diagram illustrating a hypothetical purchase history profile of
three items.
FIG. 5: illustrates one specific implementation of the sequence of steps of FIG. 2.

FIG.6: Experimental Product Taxonomy of Mobile Commerce site
FIG.7: Action performed by customer in M-commerce site and possible data collection.
FIG.8: Decision tree to playing Tennis. FIG. 9: is a recommendation system.
DESCRIPTION OF THE INVENTION
Mobile-Commerce is accepted word wide and a very popular way of doing business by using portable device. To increase the profit in business mobile-commerce companies comes with a concept of recommendation system (RS). Recommendation system provides prediction of items that may be purchased by customer. The objective of RS is to provide suggestions of items for derived variants from frequent purchases by customers. RS have three elements: Users, items and transactions. Users are people who use internet or social media platforms to purchase product. Items are products available in mobile-commerce applications. Transactions are the records of buying and selling items between seller and consumer. RS uses all the three elements for providing recommendation of product to customer.
FIG. 1: illustrates the basic components of the Amazon.com Web site 30, including the components used to implement the Recommendation Service. The arrows in FIG. 1 show the general flow of information that is used by the Recommendation Service. As illustrated by FIG. 1, the Web site 30 includes a Web server application 32 (“Web server”) which processes HTTP (Hypertext Transfer Protocol) requests received over the Internet from user computers 34. The Web server 34 accesses a database 36 of HTML (Hypertext Markup Language) content which includes product information pages and other browsable information about the various products of the catalog. The “items” that are the subject of the Recommendation Service are the titles (regardless of media format such as hardcover or paperback) that are represented within this database 36.

FIG. 2: the first step (step 80) of the recommendations-generation process involves identifying a set of items that are of known interest to the user. The “knowledge” of the user's interest can be based on explicit indications of interest (e.g., the user rated the item highly) or implicit indications of interest (e.g., the user added the item to a shopping cart). Items that are not “popular items” within the similar items table 60 can optionally be ignored during this step.
In the embodiment depicted in FIG. 1, the items of known interest are selected from one or more of the following groups: (a) items in the user's purchase history (optionally limited to those items purchased from a particular shopping cart); (b) items in the user's shopping cart (or a particular shopping cart designated by the user), (c) items rated by the user (optionally with a score that exceeds a certain threshold, such as two), and (d) items in the “recent shopping cart contents” list associated with a given user or shopping cart. In other embodiments, the items of known interest may additionally or alternatively be selected based on the viewing activities of the user. For example, the recommendations process 52 could select items that were viewed by the user for an extended period of time and/or viewed more than once. Further, the user could be prompted to select items of interest from a list of popular items.
For each item of known interest, the service retrieves the corresponding similar items list 64 from the similar items table 60 (step 82), if such a list exists. If no entries exist in the table 60 for any of the items of known interest, the process 52 may be terminated; alternatively, the process could attempt to identify additional items of interest, such as by accessing other sources of interest information.
In step 84, the similar items list 64 are optionally weighted based on information about the user's affinity for the corresponding items of known interest. For example, a similar items list 64 may be weighted heavily if the user gave the corresponding popular item a rating of “5” on a scale or 1-5, or if the user purchased multiple copies of the item. Weighting a similar items list 64 heavily has the effect of increasing the likelihood that the items in that list we be included in the

recommendations that are ultimately presented to the user. In one implementation described below, the user is presumed to have a greater affinity for recently purchased items over earlier purchased items.
The similar items list 64 are preferably weighted by multiplying the commonality index values of the list by a weighting value. The commonality index values as weighted by any applicable weighting value are referred to herein as “scores.” In other embodiments, the recommendations may be generated without weighting the similar items lists 64.
If multiple similar items list 64 are retrieved in step 82, the lists are appropriately combined (step 86), such as by merging the lists while summing the scores of like items. The resulting list is then sorted (step 88) in order of highest-to-lowest score. In step 90, the sorted list is filtered to remove unwanted items. The items removed during the filtering process may include, for example, items that have already been purchased or rated by the user, and items that fall outside any product group (such as music or books), product category (such as non-fiction), or content rating (such as PG or adult) designated by the user. The filtering step could alternatively be performed at a different stage of the process, such as during the retrieval of the similar items lists from the table 60. The result of step 90 is a list (“recommendations list”) of other items to be recommended to the user.
In step 92, one or more additional items are optionally added to the recommendations list. In one embodiment, the items added in step 92 are selected from the set of items (if any) in the user's “recent shopping cart contents” list. As an important benefit of this step, the recommendations include one or more items that the user previously considered purchasing but did not purchase. The items added in step 92 may additionally or alternatively be selected using another recommendations method, such as a content-based method.
Finally, in step 94, a list of the top M (e.g., 15) items of the recommendations list are returned to the Web server 32 (FIG. 1). The Web server incorporates this list into one or more Web pages that are returned to the user, with each recommended item

being presented as a hyper textual link to the item's product information page. The recommendations may alternatively be conveyed to the user by email, facsimile, or other transmission method. Further, the recommendations could be presented as advertisements for the recommended items.
FIG.3: illustrates the sequence of steps that are performed by the table generation process 66 to build the similar items table 60. The general form of temporary data structures that are generated during the process are shown at the right of the drawing. As will be appreciated by those skilled in the art, any of a variety of alternative methods could be used to generate the table 60.
As depicted by FIG. 3, the process initially retrieves the purchase histories for all customers (step 100). Each purchase history is in the general form of the user ID of a customer together with a list of the product IDs (ISBNs, etc.) of the items (books, CDs, videos, etc.) purchased by that customer. In embodiments which support multiple shopping carts within a given account, each shopping cart could be treated as a separate customer for purposes of generating the table. For example, if a given user (or group of users that share an account) purchased items from two different shopping carts within the same account, these purchases could be treated as the purchases of separate users.
The product IDs may be converted to title IDs during this process, or when the table 60 is later used to generate recommendations, so that different versions of an item (e.g., hardcover and paperback) are represented as a single item. This may be accomplished, for example, by using a separate database which maps product IDs to title IDs. To generate a similar items table that strongly reflects the current tastes of the community, the purchase histories retrieved in step 100 can be limited to a specific time period, such as the last six months.
In steps 102 and 104, the process generates two temporary tables 102A and 104A. The first table 102A maps individual customers to the items they purchased. The second table 104A maps items to the customers that purchased such items. To avoid the effects of “ballot stuffing,” multiple copies of the same item purchased by

a single customer are represented with a single table entry. For example, even if a single customer purchased 4000 copies of one book, the customer will be treated as having purchased only a single copy. In addition, items that were sold to an insignificant number (e.g., <15) of customers are preferably omitted or deleted from the tables 102A, 104B.
In step 106, the process identifies the items that constitute “popular” items. This may be accomplished, for example, by selecting from the item-to-customers table 104A those items that were purchased by more than a threshold number (e.g., 30) of customers. In the context of the Amazon.com Web site, to resulting set of popular items may contain hundreds of thousands or millions of items.
FIG. 5: illustrates the sequence of steps that are performed by the Instant Recommendations service to generate personal recommendations. Steps 180-194 in FIG. 5 correspond, respectively, to steps 80-94 in FIG. 2. In step 180, the process 52 identifies all popular items that have been purchased by the user (from a particular shopping cart, if designated) or rated by the user, within the last six months. In step 182, the process retrieves the similar items lists 64 for these popular items from the similar items table 60.
FIG. 9: an embodiment of a recommendation system 100 is shown that addresses the foregoing problems, among others. The recommendation system 100 includes multiple recommenders 112 for generating recommendations that target users' varied interests. The recommenders 112 provide reasons for recommending items that can be more compelling than reasons provided by other systems, thereby increasing consumer confidence in the recommendations.
The various components of the recommendation system 100 may be implemented as software applications, modules, or components on one or more computers, such as servers. While the various components are illustrated separately, they may share some or all of the same underlying logic or code.

The recommendation system 100 receives item preference data 102 and uses the item preference data 102 to produce personalized item recommendations for a target user. In an embodiment, the item preference data 102 is reflective of actions performed by the user. These actions might include, for example, purchasing items, rating items, adding items to the user's wish list, providing data on the user's friends, tagging items, searching for items, and the like. The item preference data 102 may include browse history data, purchase history data, friend’s data, tags data, and many other types of data. Some forms of item preference data 102 and their uses will be described more fully below.
The item preference data 102 is provided to a recommendation engine 110. The recommendation engine 110 includes multiple recommenders 112. In an embodiment, each recommender 112 may be implemented as a component or algorithm that generates personalized item recommendations targeted to a different interest or need of a user. The multiple recommenders 112 of the recommendation engine 110 can provide more effective recommendations than the monolithic algorithms of currently-available systems.
In an embodiment, each recommender 112 analyzes a subset of the item preference data to identify items as candidate recommendations for recommending to a user. Each recommender 112 also identifies one or more reasons for recommending the items. As discussed below, different recommenders 112 may use different types of item preference data than others to select candidate items to recommend. Different recommenders 112 may also provide different types of reasons for recommending items.
For example, a particular recommender 112 might retrieve the user's purchase history data. Using this data, the recommender 112 can find items owned by the user that are part of a series. A series might include, for instance, books in a trilogy, movies and their sequels, or all albums by a musician. If the user has purchased fewer than all the items in the series, the recommender 112 might select the remaining items as candidate recommendations and provide a reason such as, “this item is recommended because you purchased items A and B, and this item would

complete your series.” Advantageously, this reason can be more compelling than a reason such as “because you purchased items A and B, and this item is similar.” Users may therefore be more inclined to trust the reasons provided by the recommenders 112.
As another example, a recommender 112 might obtain data about a user's friends. This friend’s data might include information on the friends' birthdays, their wish lists, and their purchase histories. Using this data, a recommender 112 might suggest gifts that could be bought for a friend's upcoming birthday and provide a reason such as “this item is recommended because your friend John's birthday is on July 5th, and this item is on his wish list.” Provided with such a reason, the user might be more inclined to buy the item.
Many other examples of item preference data 102 may be used by the recommenders 112 to generate candidate recommendations and corresponding reasons. For instance, browse history data (e.g., data on user searches, clicks, and the like) may be used to provide a recommendation with the reason, “because this item is similar to an item you searched for.” Purchase history data and/or wish list data might be used to provide a recommendation with the reason, “because this item might be interesting to an early adopter such as you.” Browse history data on a browse node of interest to the user (e.g., a category browsed by the user) might be used to provide a recommendation with the reason, “because this item is a top seller in one of your favorite interest areas.” Various other forms of item preference data 102 may be used to provide recommendations with reasons such as “because you recently moved,” “because you bought an item that may need replacing,” “because most people upgrade their DVD player after two years,” or the like.
Multiple reasons may be provided by a single recommender 112, or multiple recommenders 112 may each provide the same candidate recommendation along with a different reason for that recommendation. For instance, several recommenders 112 may be used to recommend a particular war movie because 1) a user recently rated several war movies, 2) this is the bestselling movie in the war

movie category, and 3) this movie was nominated for two Academy Awards. Using multiple reasons may provide further motivation to the user to view or buy an item.
However, in certain embodiments, fewer reasons are shown to the user even when multiple reasons are available, to reduce possible information overload. In the above war movie example, the user might therefore only see the reason “because this is the bestselling movie in the war movie category.” This reason is focused and highly targeted to the user's interest of buying war movies and may be more effective than the multiple reasons provided above.
The user may also see greater diversity in the reasons that are provided. For example, the user may see one recommendation that is based on an item the user purchased, another based on one or more search queries submitted by the user, and another based on an item listed on a friend's wish list. The diversity of recommendations and reasons provided to the user may heighten user interest in the recommendations.
Advantageously, in one implementation, at least some of the
recommenders 112 are modular. Recommenders 112 can therefore be selectively
added to or removed from the recommendation engine 110. As more diverse items
or services are added to an online catalog, for instance, new recommenders 112 can
be added that target different user interests. Conversely, some
recommenders 112 may be removed from the recommendation engine 110 if they become less useful.
Some of the recommenders 112 may use particular types of behavior-based associations to select candidate items to recommend. As one example, one recommender may use purchase-based item associations, as generated by mining the purchase histories of large numbers of users, to select candidate items similar to those purchased or owned by the target user. As another example, a particular recommender may use item-viewing based associations, as generated by mining the item viewing histories of large numbers of users, to select candidate items similar to those recently viewed by the target user.

Another recommender may use behavior-based associations between particular search queries and items to select candidate items that are related to the search history of the target user. Other recommenders may select candidate items that are unusually popular in the particular geographic region of the target user, or that are unusually popular among users whose email addresses contain the same domain name (e.g., nasa.gov) as the target user. Examples of recommendation methods that use these approaches are described in the following U.S. patent documents, the disclosures of which are hereby incorporated by reference in their entirety: U.S. Pat. Nos. 6,853,982 and 6,963,850, and U.S. application Ser. No. 10/966,827, filed Oct. 15, 2004. In addition, because the recommenders 112 are modular, the recommenders 112 can be added to an existing recommendation system to improve the quality of recommendations provided by the system.
The recommenders 112 in certain implementations score the candidate recommendations. The scores can provide indications of the relative strength of the candidate recommendations. Each recommender uses one or more factors to generate the scores. As one example, a recommender 112 that provides recommendations to complete series of items owned by the user might base scores on the total number of items in a series, the number of those items owned by the user, and the sales rank of the items not owned by the user.

Claims:WE CLAIMS

1. My Invention “DYNAMIC PRODUCT TAXONOMY AND HEURISTIC APPROACH BASED RECOMMENDATION ENGINE FOR MOBILE COMMERCE “is a Recommendation systems used different filtering techniques like collaborative, content based and hybrid to provide recommendation. In This invention we use collaborative filtering technique which is also used by Netflix, Amazon and other. The collaborative recommendation systems provide predictions of user’s interest with the help of several user’s data belongs in the same set of cluster. To check user’s belongs in the same cluster, behavioral and navigational data of user’s are used. Behavioral data includes the ratio of click for a specific type of product, time used in reading the profile of product and the frequency of visits to a product exist in specific category whereas Navigational patterns includes browsing status, searching status, product click status, basket placement status, and actual purchase status of a product. Techniques for mapping item listings from a first taxonomy to a second taxonomy are described. The item listings from a first database storing a first taxonomy and item listings from a second database storing a second taxonomy are obtained. Each of the obtained item listings, a plurality of features is extracted, including at least one feature related to an image associated with the item listing and at least one feature related to text associated with the item listing. Then a mapping between item listings in the first taxonomy and item listings in the second taxonomy is created based on the plurality of features extracted by the feature extraction component. The mapping identifies which item listings in the first taxonomy correlate to a same product as which item listings in the second taxonomy. Invented implemented service analyzes purchase histories and/or other types of behavioral data of users on an aggregated basis to detect and quantify associations between particular items represented in an electronic catalog. The detected associations are stored in a mapping structure that maps items to related items, and is used to recommend items to users of the electronic catalog. The items may include products and/or categories of products.

2.According to claim1# The invention is to wherein the data repository comprises a mapping structure that associates related items. The first component uses the data values to select item pairs to include in the mapping structure.
3. According to claim1,2# The invention is to wherein the first component is configured to exclude, from the data repository, data values that do not satisfy a selected condition.
4. According to claim1,2# The invention is to wherein the input list includes at least one purchased item that was purchased by the user on a purchase date, and the second component is configured to take the purchase date into consideration in determining how much weight to give to the purchased item in generating the output list of items.
5. According to claim1,2,3# The invention is to wherein the input list includes at least one item to which the user has explicitly assigned a rating, and the second component is configured to take the rating into consideration in generating the output list of items.
6. According to claim1,2# The invention is to wherein the first component is configured to additionally take item viewing histories of users into consideration in generating the data values. The second component generates the input list of items, at least in part, by identifying items viewed by the user during browsing of an electronic catalog.
7. According to claim1,2,5# The invention is to wherein incorporating the item recommendations into a page comprises incorporating the item recommendations into a shopping cart page that displays contents of an electronic shopping cart of the user.
8. According to claim1,2,4# The invention is to wherein the item recommendations are based on said contents of the electronic shopping cart.
9. According to claim1,2# The invention is to wherein the method is performed in its entirety by a web server.
10. According to claim1,3# The invention is to wherein the recommendation service is external to the web server, and the method comprises the web server communicating with the recommendation service to obtain the item recommendations.
11. According to claim1,2# The invention is to wherein the method is performed in its entirety by a web server system that is separate from the recommendation service.
12. According to claim1,2# The invention is to further comprising generating the item recommendations with the recommendation service.
13. According to claim1,2# The invention is to wherein the item recommendations are personalized for the user based at least partly on an item ratings profile of the user.
14. According to claim1,2,5# The invention is to wherein the item recommendations are personalized for the user based at least partly on past item purchases of the user.
15. According to claim1,2,10# The invention is to wherein the item recommendations are personalized for the user based at least partly on recorded item viewing activities of the user.
16. According to claim1,2,5# The invention is to, wherein said profiling system provides profile data combinable with said weighted content relationship information relative to a characteristic attribute of the content selectable by said user of said client system.
17. According to claim1,2# The invention is to wherein said profiling information reflects explicit indications of interest and implicit indications of interest in particular instances of content.
18. According to claim1,2,6# The invention is to wherein said implicit indications of interest are derived from the selection and duration of sampling of particular instances of content.
19. According to claim1,2,9# The invention is to wherein the at least one feature related to text comprises a title.
20. According to claim1,2,12# The invention is to wherein the at least one feature related to text comprises a description.
21. According to claim1,2# The invention is to wherein the at least one feature related to the item listing, but other than a feature related to an image and a feature related to text, comprises a Universal Product Code (UPC).
22. According to claim1,2,8# The invention is to wherein the creating a mapping includes using a machine learning model to create the mapping.
23. According to claim1,2# The invention is to wherein the machine learning model is based on a random forest model.
24. According to claim1,3,8# The invention is to wherein the machine learning model is based on a logistic regression model.
25. According to claim1,3,8# The invention is to further comprising: receiving feedback from a user with regard to one or more of the item listings; and using the feedback in the machine learning model to update the machine learning model; and creating a new mapping based on the update to the machine learning model.
26. According to claim1,2,12# The invention is to including, for each of the obtained item listings, extracting at least one feature from an image associated with the item listing, wherein the creation of the mapping between item listings in the first taxonomy and item listings in the second taxonomy is also based on the at least one extracted feature from the images associated with the item listings.
27. A machine-readable medium carrying instructions, which when implemented by one or more machines, cause the one or more machines to perform operations comprising:
obtaining item listings from a first database storing a first taxonomy and item listings from a second database storing a second taxonomy; for each of the obtained item listings, extracting a plurality of features, including at least one feature related to an image associated with the item listing and at least one feature related to text associated with the item listing; and creating a mapping between item listings in the first taxonomy and item listings in the second taxonomy based on the plurality of features extracted by the feature extraction component, wherein the mapping identifies which item listings in the first taxonomy correlate to a same product as which item listings in the second taxonomy.
28. According to claim1,2,27# The invention is to further comprising performing computing at least one additional feature from the plurality of features.
29. A machine readable medium carrying machine readable instructions for controlling a machine to carry out the method.
30. According to claim1,2,27# The invention is to further comprising adjusting the number of scores combined in the combined score to reduce the impact of outliers on the normalized scores.
31. According to claim1,2,27# The invention is to wherein combining the scores for at least some of the candidate recommendations comprises summing the scores.
32. According to claim1,2,27# The invention is to further comprising assigning weights to candidate recommendations received from the first and second recommenders.
33. According to claim1,2,27# The invention is to wherein calculating normalized scores based at least in part on the combined score and the scores for at least some of the candidate recommendations comprises dividing each candidate recommendation score by the combined score.
34. According to claim1,2,28# The invention is to wherein the moving average comprises an exponential moving average.

Dated: 10/5/2020
GUJARAT UNIVERSITY
DR. SHANTI VERMA
DR. KALYANI PATEL (ASSISTANT PROFESSOR)
, Description:FIELD OF THE INVENTION
The invention “DYNAMIC PRODUCT TAXONOMY AND HEURISTIC APPROACH BASED RECOMMENDATION ENGINE FOR MOBILE COMMERCE “is relate to techniques for mapping products between different taxonomies, information filtering and recommendation systems.