DESCRIPTION
In the field of online advertising/publishing of the consumer goods and services on a comparison shopping engine, a method and system are disclosed for the publication of product information from a merchant website on a comparison shopping engine.
DEFINITIONS
Online shopping: Online shopping is purchasing of products or services over the Internet.
Merchant Website: A merchant website is a portal maintained by a merchant to publish information related to her products and to sell those products online. A merchant website may sell various products offered by the merchant. Each product may have a separate product page.
Comparison Shopping Engine (CSE): A Comparison shopping engine is a portal which allows consumers to view comparative and taxonomical information for specific products, from one or more merchants, at one place. CSE publishes product information related to various products enabling consumers to make prudent choices while purchasing products.
Product information: Product information is the information about a product. The information may include product specific fields like product name, product status, product image, product identifier such as manufacturer part number, ISBN and the like, product description and the like. The product information may also include merchant specific fields such as product price, availability in stock, shipping details, product promotion such as sale, rebate, discount and the like. The product information may further include URL of the product page on merchant website, product manufacturer related information, classification of the product on merchant website, and relevant keywords. For example, keywords for a product 'camera' may include 'optical zoom ', 'color' and 'pixel size'.

Product Data Feed: A product data feed is a file comprising at least some portion of product information. The portion of product information to be included in the product data feed may depend on comparison shopping engine requirements. The product data feed has a plurality of fields for providing product information. The product data feeds are sent to online shopping portals such as comparison shopping engines for the publication of product information.
Communication network; Communication network is a network of communication devices and stations having wired or wireless interconnection for establishing communication. Communication includes, but is not limited to, transmitting and receiving signals. Communication further includes, but is not limited to, transferring information and data such as voice, audio, video, graphics and the like. Communication network includes, but is not limited to, internet, intranet, extranet, Wide Area Network (WAN), wireless WAN (WWAN), Local Area Network (LAN), wireless LAN (WLAN), transducer links such as those using Modulator-Demodulators (modems), telecommunication network, personal area network and Global Navigation Satellite System (GNSS). Telecommunication network includes, but is not limited to, Public Switched Telephone Network (PSTN), Global System for Mobile Communications (GSM), and Code Division Multiple Access network (CDMA). Personal area network includes, but is not limited to, Bluetooth and Infrared, and Global Navigation Satellite System (GNSS).
BACKGROUND ART
In era of online shopping, consumer products are sold on a merchant's website and the products are subsequently shipped to consumers. The drawback of such an approach is that the consumers, while purchasing a product from a particular merchant website, may be unaware of other merchants offering similar products. The consumers may therefore miss out on various promotional offers and discounts from other merchants. Thus there was a need for a portal wherein a consumer, looking for a particular product, may browse through similar products from different merchants at one place. Comparison shopping engines cater to this need

and provide comparative product information from various merchants for any given product.
For publishing product information on a CSE, the merchant, who is selling the product, has to provide the product information to the CSE. The merchants provide the product information to the CSE through a product data feed. Typically, merchants create the product data feed and then send the product data feed for publication on a CSE.
The process of product data feed creation and publication have certain problems. One, the merchants generally create the product data feed manually. In a product data feed, information about every field is desirable. However, during manual creation, one or more fields may get skipped. Secondly, merchants may provide misleading information in order to achieve a certain classification of a product. The above mentioned problems may result in wrong classification and thus, improper placement of the product on a CSE.
Existing art attempts to solve the above mentioned problem to a certain extent. There exists channel management tools to manage product data feed publication process. For example 'Feedpoint' from 'Quigo', 'ShoppingAdvisor' from 'ChannelAdvisor Inc.', 'SellCore' from 'Channelintelligence' etc. Patent No. EP1759279A2 discloses a method for automated data feed creation system. Existing art therefore provide automatic data feed creation and the classification according to the comparison shopping engines based on the input provided by the merchants.
However, the above mentioned prior art do not provide a solution when the information in the product data feed is incomplete or misleading.
Therefore, there is a need of an improved automated system for product data feed creation and its publication on the comparison shopping engine. Further, there is a need for a system to regularly monitor the published data feeds so as to capture and therefore prevent any wrong classification of products.

BRIEF DESCRIPTION OF DRAWINGS
FIG. 1 is a schematic illustrating an environment of a product data feed publication on a comparison shopping engine according to an embodiment of the invention;
FIG. 2 is a block diagram illustrating a system of product data feed publication on a comparison shopping engine according to an embodiment of the invention;
FIG. 3 is a flow diagram illustrating a method for publishing the product data feed on the comparison shopping engine according to an embodiment of the invention;
FIG. 4 is a flow diagram illustrating the step of aligning a product data feed according to an embodiment of the invention.
DESCRIPTION OF EMBODIMENTS
In the following description, specific details are set forth in order to provide a thorough understanding of the invention. However, it wilt be apparent to a person ordinarily skilled in the art, that the invention may be practiced without these specific details. Various aspects and features of example embodiments of the invention are described in detail hereinafter.
FIG. 1 is a schematic depicting an environment of product data feed publication on a comparison shopping engine. According to an embodiment of the invention, the environment comprises a merchant website 100, a communication network 102, a data feed publisher 104 and a comparison shopping engine 106. Merchant website 100 may comprise a plurality of product pages; a product page may comprise product information of a product from the corresponding merchant. The product information is characteristics of the product such as product name, type etc and the merchant specific information such as product condition, product price etc. The product information is sent to comparison shopping engine 106 in the form of a product data feed. Thus each product data feed is associated with a product and also with a corresponding product page on merchant website 100. A product data

feed comprises a product information organized in a plurality of fields such as product name, product condition, image, price offered by the merchant etc. The product data feed is sent for publication on comparison shopping engine 106. Consumers visiting comparison shopping engine 106 using for example, a web browser, may view the published product data feed and, if interested, may click on link provided. Thereafter, consumers are directed to the product page, associated with the product data feed on merchant website 100.
Data feed publisher 104 is an automated system for product data feed generation and publication. According to some embodiments of the invention, data feed publisher 104 resides on a computer and provides a user interface. Data feed publisher 104 has various modules which may be written in commonly known programming languages like JAVA/C/C++, and scripting languages like PERL. A user may interact with data feed publisher 104 through the user interface. Data feed publisher 104 connects to merchant website 100 via communication network 102. Data feed publisher 104 collects product information of the product such as product name, product condition, image, price offered by the merchant etc from merchant website 100, creates a product data feed corresponding to the product information and publishes the product data feed on comparison shopping engine 106 via communication network 102. Data feed publisher 104 also audits the published data feeds for any inaccuracy in classification.
Although the invention is discussed in conjunction with one merchant website and one comparison shopping engine, it will be apparent to any person skilled in the art that data feed publisher 104 may cater to the needs of several merchant websites and several CSEs.
FIG. 2 is a block diagram illustrating data feed publisher 104 according to an embodiment of the invention. Data feed publisher 104 comprises a CSE & Product URL Input Module 200, a data extraction module 202, database 204, a data feed generator module 206, an uploader 208 and an auditor 210.

CSE & Product URL input module 200 may receive an input data for a product. The input data comprises name of comparison shopping engine 106 where product information of a product has to be published, the product name and a URL of product page corresponding to the product on merchant website 100. According to some embodiments of the invention, a user may provide input data for a product to CSE & Product URL input module 200 through the user interface. According to some embodiments, CSE & Product URL input module 200 may retrieve the input data from a database 204. It will be apparent to a person skilled in the art that the means for receiving input data is not a limitation of the invention. The CSE & Product URL input module 200 then provides the input data to data extraction module 202.
Data extraction module 202 connects to merchant website 100 through communication network 102. In one embodiment of the invention, data extraction module 202 takes the URL of the product page provided as input, uses a web crawler to visit the corresponding product page and extract the product information such as product name, product condition, image, price offered by the merchant etc. The web crawler may be computer software written in PERL script. Data extraction module 202 stores the extracted product information in database 204.
Database 204 is a collection of data regarding products on the merchant website 100. Database 204 organizes the product information related to the each product separately. For each product, database 204 stores input data, product information extracted, and secondary product pages information. Secondary product pages information for a product includes manufacturer website URL for the product, other related sources of product information including other merchant website URLs for the product etc. manufacturer website URL is the URL of the website of the manufacturer who has manufactured the product. Other merchant website URL may be URLs of the websites which are relevant to the product. For example, the product may be a mobile phone having a specific camera. Websites of other merchants providing detailed information regarding the mobile phone may constitute other merchant website URLs. Further, a website having information on the specific camera may also be one of the other merchant website URL.

Database 204 also stores a merchant taxonomy of products for each merchant. Merchant taxonomy may comprise the classification of different products into different categories by the merchant.
Further, database 204 may also store certain data for each comparison shopping engine. The data regarding comparison shopping engine 106 comprises taxonomy information for comparison shopping engine 106, alignment rules for comparison shopping engine 106, and the like. Taxonomy of comparison shopping engine 106 may be different from the merchant taxonomy. Taxonomy of comparison shopping engine 106 is discussed in conjunction with FIG. 3. Alignment rules for a CSE are discussed in detail in conjunction with FIG. 4.
Taxonomy contains a plurality of categories. Each category may include one or more products based on the type of product. For example, on a website all consumer products may be categorized into 20 different categories. Example of categories may be 'Notebooks and PDAs' and 'Mobile phones and accessories'. A 'Dell Inspiron' is a laptop computer that may be categorized under the category 'Notebooks and PDAs'. Similarly 'Nokia 6600' is a cell phone which may be classified under the category 'Mobile phone and accessories'. Taxonomy information includes names of all categories in the taxonomy. Taxonomy may be hierarchical. The taxonomy information further includes synonyms for the category name. For example, a category 'office wear shirts' on comparison shopping engine 'A' may be called as 'formal shirts' on other comparison shopping engine 'B'. Taxonomy information also includes keywords related to each category. For example, a category 'formal shirt' may have keywords such as cufflink, collar, sleeves, color and the like. Other information such as the product data feed format, field names, synonyms used for the field names, and units used on comparison shopping engine 106 may be stored in database 204.
Data feed generator module 206 comprises of a data feed creator 212, a classifier 214 and an aligner 216. Data feed generator module 206 generates a product data feed for the product using the information stored in database 204. In the process of data feed generation, data feed generator module 206 creates a product

data feed, classifies the product data feed into one or more categories and aligns the product data feed such that the product data feed contains accurate and complete product information. Each step is discussed in detail subsequently.
Data feed creator 212 creates the product data feed using information in database 204. Data feed creator 212 forwards the product data feed to classifier 214.
Classifier 214 is a classification module which classifies the product data feed into one or more categories by natural data processing of the product information. The one or more categories are identified according to the taxonomy information of comparison shopping engine 106 stored in database 204. Classifier 214 may update product information stored in database 204 by including the one or more identified categories in the product information.
In an embodiment of the invention, the natural data processing of the product information may be a machine learning technique of the classification of objects. For example, a Naive Bayes classification technique of supervised machine learning may be used for the classification of the products. A Plurality of keywords is identified based on natural language processing. Identifying the keywords and machine learning calculations on the identified keywords lead to the classification of the product data feed in one or more categories.
Aligner 216 receives the product data feed from classifier 214 and aligns all the information contained in the product data feed according to the comparison shopping engine 106. In an embodiment of the invention, aligner 216 verifies the data feed for the completeness of all the fields. All the incomplete information is filled in aligner 216 using data from database 204. The values of the fields in the product data feed may be normalized using database 204. For example, the units of 'length' in the product data feed may be 'centimeters' whereas the CSE required the length' in 'inches'. So aligner will modify the product data feed changing the units of 'length' to Inches'.

In case the produu UQlQ .ecu oumanis a Keyworas' field, aligner 216 generates a set of keywords associated with the product from the product information stored in database 204. Aligner 216 also verifies all the values of the fields in the product data feed. Verification is described in conjunction with FIG. 4. Aligner 216 provides a classified and aligned product data feed to uploader 208.
Uploader 208 is a module for the publishing of the product data feed on comparison shopping engine 106. Uploader 208 is connected through communication network 102 to comparison shopping engine 106. In an embodiment of the invention, uploader 208 receives the product data feed from data feed generator module 206 and sends the product data feed to comparison shopping engine 106 via communication network 102.
Auditor 210 is a module which checks the status of the published product data feed on comparison shopping engine 106. Auditor 210 may crawls through the comparison shopping engine 106 and obtains status information of the product. The status information includes the name of the one or more categories, where the product data feed is categorized on comparison shopping engine 106, the URL of the product page on the merchant website etc. Auditor verifies the status information by matching with respect to information stored in database 204.
In case of any mismatch between the status information and the information stored in database, auditor 210 gives a signal to data feed generator module 206 to follow the process of data feed generation and re - publication of the product data feed.
FIG. 3 is a flow diagram illustrating a method for product data feed generation and publication on comparison shopping engine 106.
At step 300, an input data for a product is provided to data feed publisher 104. The input data comprises name of comparison shopping engine 106, the product name and the corresponding product page URL. In an embodiment of the invention, the input data may be provided in the form of spreadsheets or flat files

containing the name of comparison shopping engine 106, the product name and the corresponding product page URL. In another embodiment, the input can be provided by filling an html form, the html form requesting information for name of comparison shopping engine 106, the product name and the corresponding product page URL. For example, the merchant enters 'Shopzilla' as comparison shopping engine 106, 'ABC Camera' as product name and the product page URL in the form. The CSE & Product URL input module 200 receives the input data and provides the input data to data extraction module 202.
In one embodiment, the product page URL may not be provided as an input to data feed publisher 104. Instead, the merchant website URL may be provided and data extraction module 202 extracts the product page URL, while crawling over merchant website 100, on basis of matching of product name with product page title.
In another embodiment of the invention, the input can be provided by an automated system, which periodically crawls merchant website 100 and collects information regarding new products and corresponding product pages. Step 300 is performed by CSE & Product URL input module 200, which provides the input data to data extraction module 202.
At Step 302, data extraction is performed by data extraction module 202. Data extraction is a process of collecting the relevant information regarding a product regarding product name, product status, product image, product identifier such as manufacturer part number or ISBN, product description and merchant specific information such as price, promotion etc from merchant website 100. In one embodiment, information between HTML or XML tags on the product page may be extracted using PERL script based programs. The information may be extracted from title of the product page, metadata and body of the product page.
For example, a product page may present 'formal shirts' for men. The page may present information on price, colors, collar type and the like in tabular format or descriptive format between the 'body' tags. The extraction module captures

all the information between the 'body' tags. Similarly other relevant information between other tags is also collected. In another embodiment of the invention, the information on the product page may be in other formats such as XML. The information is collected by data extraction module 202 and stored in database 204.
At step 304, the product data feed is created by data feed creator 212 using the product information stored in database 204. In some embodiments of the invention, values for all the relevant fields or attributes of the product data feed are identified in the product information. In an embodiment of the invention, an attribute set is maintained corresponding to each attribute. The attribute set may be constructed based on requirements from various comparison shopping engines. For example, an attribute set corresponding to attribute 'Warranty' may comprises of 'Warranty', 'After Sale Services Warranty', 'Warranty on Parts' and the like. In identifying value for a particular attribute, data corresponding to items in the attribute set is searched in product information. The data searched forms the value for the corresponding attribute.
For example, attribute set for attribute 'price' may be '$', 'INR', and 'price'. A search may be performed for '$' in the product information. As an example product information may read $25, Data obtained in vicinity of '$', i.e. 25, may be used as value for attribute 'price'. In case there are more than one '$' sign, other items in the attribute set may be searched to come up with accurate value for 'price'.
Values are specified for each attribute of the product data feed. Thus created product data feed includes product specific fields and corresponding vaiues. Product specific fields include product name, product status, product image, product identifiers, product description and the like. The product data feed may also include merchant specific fields like product price, availability in stock, shipping details, product promotion and the like.
In another embodiment of the invention, the information related to the product may be supplied by the merchant manually or by some system, then the creation of product data feed is followed by the verification of the information from

the information extracted from merchant website 100 and stored in database 204. The verification is done by data feed creator 212.
At step 306, classification of the product data feed is performed. The classification process comprises the identification of one or more categories to which the product may belong. The categories of the product data feed are identified from the categories in the taxonomy defined by comparison shopping engine 106. For example, 'Laptops and notebooks' may be a category of the products, which contains products such as laptops, notebooks, PDAs etc. Any product, which falls under those products, will be classified under 'Laptops and notebooks'. Data feed publisher 206 may be connected with multiple comparison shopping engines. In this case, classifier 214 maintains a taxonomy which contains all the categories of all comparison shopping engines. For example, suppose classifier 214 maintains a taxonomy corresponding to two comparison shopping engines. Then the number of categories in the taxonomy of classifier 210 will be a superset of categories of both comparison shopping engines. In this case, if a product is classified as 'formal shirt' in first comparison shopping engine and 'office wear shirt' in second comparison shopping engine, then classifier 214 maintains both categories as two separate categories and the product data feed will be classified in both the categories. In another embodiment of the invention, classifier 214 may maintain a single category X corresponding to both the categories i.e. 'formal shirt' and 'office wear shirt' and maintains a list of synonyms of name of the category 'X'. The list will then include both 'formal shirt' and 'office wear shirt'.
Classifier 214 performs natural data processing of the product information to categorize the product. Supervised machine learning techniques such as Naive Bayes Classification, maximum entropy algorithm etc. or their modifications may be applied for the classification of the product data feed. Keywords are identified using the natural language processing and statistical calculations on the keywords based on above mentioned algorithms are used to calculate relevance of the product to every category in the taxonomy.

According to some embodiments of the invention, classifier 214 may crawl through the product page and perform a natural language processing of the product page on merchant website 100. Classification may thus be done based on natural language processing of the product page. Classifier 214 may update the product information in database 204 by including the one or more identified categories for the product.
At step 308, the process of alignment of the product data feed is performed by aligner 216. The process of alignment of the product data feed comprises normalizing the field names and the corresponding values. The normalization process comprises formatting the data feed according to comparison shopping engine 106 requirements. The process of alignment further comprises of filling the blank fields of the data feed using the secondary product page information. The URLs stored for secondary product pages in database 204 are used to obtain the information regarding the product. The values of fields may be obtained from information extracted from secondary product pages. According to some embodiments of the invention, information is extracted through natural language processing of the secondary product pages. The embodiment includes crawling through the secondary product pages and collecting information regarding the product using natural language processing. The attribute sets of fields may assist in extracting information from the secondary product pages. The process of alignment also comprises of developing the keywords for the product and remapping or validation of the fields in the product data feed. Aligner 308 may update the product information in database 204 by including the revised information for various data fields.
After the process of alignment, data feed generator module 206 sends the product data feed to uploader 208.
At step 310, uploader 208 uploads the product data feed on comparison shopping engine 106. Uploader 208 inputs the product data feed from data feed generator module 206 and using communication network 102, publishes the product data feed on comparison shopping engine 106 using communication

network 102. In one embodiment of the invention, uploader 208 uses standard internet send/receive commands such as ftp push for the uploading of the data feed. In other embodiment, the product data feed can be uploaded to comparison shopping engine 106 by communication network 102 such as telecommunication network.
After the product data feed is published on comparison shopping engine 106, auditor 210 may perform periodic checks to ensure appropriate categorization of the product.
At step 312, auditor 210 performs auditing of the product data feed published on comparison shopping engine 106. In an embodiment of the invention, auditor 210 crawls through comparison shopping engine 106 to audit the product data feed. Auditor 210 checks whether the product has been placed in comparison shopping engine 106. Auditor 210 further checks the category where the product has been placed. The auditor may also check the URL of the product page on merchant website 100 provided in the product data feed. Auditor 210 may match the category and the URL of the product page with the product information stored in database 204. Auditor 210 decides state of the product data feed based on the checks. If the state is not found OK, then auditor 210 sends a signal to data feed generator module 206 to repeat the process of product data feed generation and re-publication of the product data feed and step 304 is executed.
FIG. 4 is a flow diagram elaborating the step of aligning the product data feed in the process of data feed generation. Aligner 216 performs the process of alignment.
At step 400, normalization of the fields and the corresponding values of the product data feed is performed. Normalization comprises changing the field names in the product data feed as per required in comparison shopping engine 106 and changing the units and values of the fields corresponding to comparison shopping engine 106. In one embodiment of the invention, all field names, the corresponding synonyms, all the corresponding values, variants of values and rules

for the normalization are stored in database 204. For example, merchant website may refer to a field as 'Product Status'; but comparison shopping engine 106 may require a field named Product Condition'..Aligner 216 will check database 204 for the synonyms for the 'Product Status'. If 'Product Condition' is found in the synonym list of the above mentioned field, then field name is changed to 'Product Condition' in the product data feed. A user may store the synonyms in database 204 through a user interface. In another example, the value for a field is given from the set {new, used}, while comparison shopping engine 106 requires value for the field from the set {new, second-handed}. The aligner 216 then may change the field {new, used} to {new, second-handed}, if applicable.
At step 402, blank fields of the product data feed are populated. In an embodiment of the invention, if some fields of the product data feed are left blank because of insufficient information in the product information extracted from merchant website 100 by data extraction module 202, then those fields may be filled from the data captured using the second product pages information. In another embodiment, the product specific information may be taken from the second product page such as other merchant websites as the product specific information is likely to be similar.
At step 404, the process of developing keywords for the product data feed is performed by aligner 216. The keywords (such as 'zoom', 'pixel' etc for a 'camera') are taken from database 204. The keywords may increase the occurrence of the product in the search results on comparison shopping engine 106. The keywords are extracted from the various sources including secondary product page information and the keywords related to a category stored in database 204. For example, other merchant website lists 'zoom' in a 'Nikon' product which merchant website 100 doesn't, then 'zoom' may be used for as keyword for the product.
At step 406, remapping or validation of the fields of the product data feed is performed. Aligner 216 checks for consistency for the fields and their values of the product data feed stored in database 204. For example, a TV set cannot have a price below $10 and a pair of socks cannot have price as high as $1000. Other

examples include: price of a refurbished item cannot be more than price of a new item; shipping charges should not be more than price of the product etc. Some category specific rules may be applied, for example, for 'SLR camera' 'view finder' field value should be 'No'. This is apparent to the person skilled in the art that several other consistency checks may be performed based on predefined rules. Exceptions may be allowed after validations.
It is understood that while certain forms of this invention have been illustrated and discussed, it is not limited thereto. However, various alterations may be made thereto without departing from the broader spirit and scope of the invention. The specification and drawings are, accordingly, to be regarded in an illustrative sense rather than a restrictive sense.

WE CLAIM:
1. A method of publishing a product data feed on a comparison shopping
engine, the product data feed being associated with a product page, the
method comprising the steps of:
b) Creating the product data feed;
c) Classifying the product data feed into one or more categories;
d) Aligning the product data feed based on the comparison shopping engine requirements, wherein the step of aligning of the product data feed is performed based on natural language processing of secondary product pages; and,
e) Sending the product data feed for publication on the comparison shopping engine.

2. The method of claim 1 further comprises the step of auditing the product data feed published on the comparison shopping engine,
3. The method of claim 2 wherein auditing comprises checking the product data feed for accuracy of placement and correctness of information,
4. The method of claim 1 wherein the step of classification further comprises:
a. Extracting product information from the product page, wherein
extraction is done through crawling the product page;
b. Classifying the product data feed based on natural language
processing of product information.
5. A method of publishing a product data feed on a comparison shopping
engine, the product data feed being associated with a product page, the
method comprising the steps of.
a) Creating the product data feed, wherein creation comprises the
steps of:
i. Extracting product information from the product page, wherein
extraction is done through crawling the product page;
ii. Creating the product data feed using a comparison shopping engine
Information and the product information;

b) Classifying the product data feed into one or more categories, wherein the classification of the product data feed comprises natural language processing of the product information; and
c) Sending the product data feed for publication on the comparison shopping engine.

6. The method of claim 5 further comprises the step of aligning the product data feed.
7. The method of claim 6, wherein the step of aligning further comphses the step of obtaining missing information related to product data feed by crawling through a second product page.
8. The method of claim 6 , wherein the step of aligning further comprises the steps of:

a) Normalizing data fields of the product data feed and corresponding values of the data fields;
b) Populating the blank data fields of the product data feed;
c) Developing keywords for the product data feed; and
d) Remapping the data fields of the product data feed,

9. The method of claim 8, wherein the step of populating the blank data fields further comprises the step of obtaining information related to blank data fields by crawling through a second product page.
10. The method of claim 5 wherein the categories are defined by the comparison shopping engine.
11.A system for publishing a product data feed on a comparison shopping engine, the product data feed being associated with a product page, the system comprising:
a) A data extraction module for extracting product information from the product page;
b) A data feed generator module for generating the product data feed from the product information;
c) An uploader for sending the product data feed for publication on the comparison shopping engine; and
12.The system of claim 11 further comprising an auditor for auditing the published product data feed.

13. The system of claim 11 further comprising a database for storing the extracted
product information. 14.The system of claim 11, wherein the data feed generator module further
comprising:
a) A data feed creator for creating the product data feed from the
product information;
b) A classifier for classifying the product data feed into one or more
categories, wherein the classification of the product data feed
comprises natural language processing of the product page;
c) An aligner for aligning the product data feed based on the
comparison shopping engine.
15.The system of claim 11, wherein the aligner obtains missing information related to product data feed by crawling through a second product page.