FORM 2
THE PATENTS ACT, 1970
(39 of 1970)
&
THE PATENT RULES, 2003
COMPLETE SPECIFICATION
(See Section 10 and Rule 13)
Title of invention:
INTELLIGENT SYSTEM AND METHOD TO PROVIDE RECOGNITION AND EXTRACTION OF AN INFORMATIVE SEGMENT
Applicant
TATA CONSULTANCY SERVICES LIMITED A Company Incorporated in India under The Companies Act, 1956
Having address:
Nirmal Building, 9th Floor,
Nariman Point, Mumbai 400021,
Maharashtra, India
The following specification particularly describes the invention and the manner in which it is to be performed.

FIELD OF THE INVENTION
[001] The present invention in particular relates to a system and method for processing data objects. More particularly, the present invention relates to a system and method for identifying and extracting an informative segment from the data object.
BACKGROUND OF THE INVENTION
[002] Technology has become advanced with respect to searching an informative portion in a particular document. In day to day life, as data storage in electronic form has become a common practice, the technology has developed lot of advance techniques in order to serve data storage in almost every format. Now days, even the hand written documents, images, soft copy of receipt etc could be easily scanned and stored. Although, fast development in storing methods is there but to pull out required part of information from few of such documents is still not easy. Constraints are there when one needs to reach a particular portion in the stored data in Image format, which is a non-searchable document.
[003] In most of the service/product providing sectors like BPO's, call centers, government offices (like passport office, license office, etc) certain hand written and scanned copies of documents (mortgage applications, insurance claims, tax returns etc) is stored. For these sectors, the documents form the backbone of daily operations and most of these are from different sources like customer, business partner, vendor, government, semi-government agencies due to which more often these are unstructured and their format depends on the source from where the document emerge. Such type of data is stored in mass and when a need is there to search some information from the same data, one has to locate and extract that segment manually as these are Images, which are not searchable. It's a time consuming task and a hard nut to crack to search for just a part of information from large amount of data. One has to scroll down a page and manually locate for the required content to be referred and/or to be validated.

[004] The most frequently used method uses the OCR/IVR technique to convert the scanned images of handwritten, typed or printed data into an electronic or mechanical form so that automated data entry or data review can be enabled from the relevant document. Although, this technique is quite useful but it is very difficult to ensure over the accuracy of the results thus fetched in a case where the data is not structured or in a case where a noise has entered after the format is changed.
[005] Therefore, there is a need of such a solution which is capable of providing an advanced search as per the user's requirement while ensuring the accuracy of the information thus retrieved.
OBJECTS OF THE INVENTION
[006] The primary object of the invention is to facilitate an intelligent system and
intelligent method for recognition and extraction of an informative segment from an
object data. [007J The another object of the invention is to provide a user interface to configure a
user specific search query with respect to extraction of the informative segment from
the object data. [008] The another object of the invention is to provide a mapping module to map the
search query with the data stored in different data stores to ensure over the accuracy
ofresults. [009] Yet another object of the invention is to provide a validation module to further
ensure accuracy by referring to the object data stored in an original format.
SUMMARY OF THE INVENTION
[0010] The present invention provides an intelligent system facilitating recognition and extraction of an informative segment from an object data. The intelligent system comprises of a user interface configured to define a customized search query by receiving one or more input parameters from a user, a data conversion module configured to convert original format of said object data into a machine readable

format and store the machine readable format in a respective data store in order to search and recognize the informative segment from the object data and a processing engine by using a set of programmed instructions configured to retrieve one or more close results of the informative segments with respect to a particular query by searching in one or more data store. The processing engine further comprises of a data mapping module configured to map the input parameters of the customized search query with the data stored in the respective database by using a text manipulation methodology and further parse through said data object in order to recognize the informative segment and a validation module configured to check an accuracy of the informative segment thus recognized by referring to the object data in the original format. The intelligent system further comprises of an output generation module configured to retrieve one or more close results of validated informative segments of object data with respect to the user's search query. [0011] The present invention also provides an intelligent method facilitating recognition and extraction of an informative segment from an object data. The intelligent method comprises of defining a customized search query by receiving one or more input parameters from a user in order to search and recognize the informative segment from the object data, converting original format of said object data into a machine readable format and store the machine readable format in a respective data store and processing said search query by using a set pf programmed instructions to retrieve one or more close results by searching in one or more data store. The processing further comprises of mapping the input parameters of the customized search query with that of the data stored in the respective data store by using text manipulation methodology and further parsing through said data object in order to recognize and extract the informative segment. The method further comprises of checking an accuracy of the informative segment thus recognized by referring to the data object in the original format and retrieving one or more close results of validated informative segments of object data with respect to the user's search query.

BRIEF DESCRIPTION OF DRAWINGS
[0012] Figure I illustrates system architecture for data recognition and extraction in
accordance with an embodiment of the invention. [0013] Figure 2 illustrates a flow chart towards the intelligent method facilitating
recognition and extraction of an informative segment from an object data in
accordance with an embodiment of the invention.
DETAILED DESCRIPTION OF THE INVENTION
[0014] Some embodiments of this invention, illustrating its features, will now be discussed:
[0015] The words "comprising", "having", "containing", and "including", and other forms thereof, are intended to be equivalent in meaning and be open ended in that an item or items following any one of these words is not meant to be an exhaustive listing of such item or items, or meant to be limited to only the listed item or items.
[0016] It must also be noted that as used herein and in the appended claims,'the singular forms "a", "an", and "the" include plural references unless the context clearly dictates otherwise. Although any systems, methods, apparatuses, and devices similar or equivalent to those described herein can be used in the practice or testing of embodiments of the present invention, the preferred, systems and parts are now described. In the following description for the purpose of explanation and understanding reference has been made to numerous embodiments for which the intent is not to limit the scope of the invention.
[0017] One or more components of the invention are described as module for the understanding of the specification. For example, a module may include self-contained component in a hardware circuit comprising of logical gate, semiconductor device, integrated circuits or any other discrete component. The module may also be a part of any software programme executed by any hardware entity for example processor. The implementation of module as a software programme may include a set of logical instructions to be executed by the processor or any other hardware entity. Further a

module may be incorporated with the set of instructions or a programme by means of an interface.
[0018] The disclosed embodiments are merely exemplary of the invention, which may be embodied in various forms.
[0019] The present invention relates to an intelligent system and intelligent method that facilitates recognition and extraction of an informative segment from an object data. Once the object data in original format (in mass) is stored in a data store, the same may be searched for extracting or pulling a piece of information from a particular data object after changing the format of data into a searchable one. To make this process more efficient, the intelligent system and intelligent method facilitates in defining a customized search query by receiving one or more input parameters from a user in order to search and recognize an informative segment from the data object. The data object is converted into a machine readable format and stored in a data store. Further, the search query is processed to retrieve one or more close results by searching in one or more data stores. The results thus searched are also validated by referring to an original source of data (i.e. object data in the original format).
[0020] In accordance with an embodiment, referring to figure 1, the intelligent system (100) comprises of a user interface (102) is configured to define the customized search query by receiving one or more input parameters from the user. The user interface (102) may include keyboard, touch pad or some similar hardware entity.
[0021] The intelligent system (100) further comprises of a data conversion module (104) configured to convert an original format of said object data into a machine readable format. The intelligent system (100) further comprises of a processing engine (106) configured to retrieve one or more close results with respect to a particular query by searching in one or more data store. The processing engine (106) further comprises of a data mapping module (108) and a validation module (110). Also, the intelligent system (100) comprises of an output generation module (112) which is configured to retrieve one or more validated close segments of object results with respect to the user's customized search query.

[0022] In accordance with an embodiment, referring to figure 1, the data object in original format is first stored in a particular data store. For example, soft copy of a user's identity proof (pan card, voter id etc) is scanned and then stored in the data store. Now that document (data object) is required to be searched in order to extract an informative segment (say permanent account number of the user) from that data object.
[0023] In order to overcome the manual burden, the intelligent system (100) of the present invention provides the user interface (102) configured to define the customized search query by receiving one or more input parameters from the user. These input parameters may be re-defined or customized as per a user's requirement (as shown in step 202 of figure 2). These input parameters by using a set of programmed instructions are further used to automatically execute the search query to pull out the informative segment from the data object stored in the data store.
[0024] Once the search query is defined by providing one or more input parameters, the next step will be to process the query. While searching, the format of data is a big constraint as many a times the search may not be initiated due to unstructured data format. For the same, the intelligent system (100) thus provides the data conversion module (104) configured to convert the original format of the data object into the machine readable format (as shown in step 204 of figure 2). The conversion of the data object from the original format into a machine readable format is implemented by an Optical Character Recognition technique. The converted data object is stored in . a respective data store in order to search and recognize the informative segment from the data object.
[0025] The data object further comprises of scanned images (in tiff, jpg), which may include machine printed text and combination of machine printed text and hand written data. The informative segment further comprises of a text part of the data object. The original format of data further comprises of a structured data, an unstructured data, a semi-structured data or a combination thereof.
[0026] Since the data object is stored in mass in plurality of data stores, data object with some similar keywords may be found in more than one data stores. It is thus very

important to search in almost all such data stores where there is a probability of obtaining similar search results. This will ensure the improved accuracy in search results with respect to the particular search query. In order to serve the purpose of searching in more than one data stores for one particular query, the intelligent system (100) is provided with an integration module (114) (as shown in figure 1).
[0027] Still referring to figure 1. the search query is then passed to the processing engine (106) configured to retrieve one or more close results of informative segment (as shown in step 206 of figure 2) with respect to the customized search query by searching in one or more data store by enabling the communicating with the help of the integration module (114).
[0028] The at least one processor engine (106) may be implemented as one or more microprocessors, microcomputers, microcontrollers, digital signal processors, central processing units, state machines, logic circuitries, and/or any devices that manipulate signals based on operational instructions. Among other capabilities, the at least one processor engine (106) is configured to fetch and execute computer-readable instructions stored in its memory.
[0029] The integration module (114) connects through various data stores and data repositories in order to retrieve the results with respect to the informative segment thus desired. The processing engine (106) implements the set of programmed instructions in order to retrieve close results. Set of programmed instructions is a framework that includes the defining the metadata like:
[0030] Document Classification Information - Required input Image type like tiff, jpg etc. Document characteristics like Machine printed, Handwritten etc
[0031] Document Repository Information - Repository information like File Net, Documentum etc with server details to connect to the repository. Connection to repository to be achieved using standard interfaces like Web Services, Repository specific API(s). On successfully connecting to document repository in a non-interactive mode, based on the metadata defined and made available based on the context of the processing, one or more relevant data stores are accessed and parsed non-interactively.

[0032] Parsing Catalog Information - Within each of the defined document the required search criteria description with standard verbiages associated and additional secondary verbiages that can be used. These verbiages are keywords which are parsed by the system (] 00) from the input data object to render the required snippet.
[0033] Process Mapping Information - Define Operations Processes, Sub process and mapping of the documents required with the Search Criteria's for the same
[0034] The processing engine (106) further comprises of the data mapping module (108) configured to map the input parameters of the search query with the data stored in the respective database by using a text manipulation methodology (as shown in step 208 of figure 2) and further parse through said data object in order to recognize the informative segment. This step of parsing is performed by means of a parsing module (not shown in figure) communicatively coupled with the processing engine (106). Within each of the data object, the data is parsed by way of standard verbiages/secondary verbiages and rendering required snippets. For example, say for a specific business process requires an Application document Personal Details, Contact details, Employment Details section to be viewed then for each of the required section the associated keyword/verbiages like "Personal Information", "Current Contacts" /'Employment Information" which are specific to the required contents in the input data object are included in metadata definition.
[0035] By way of an exemplary embodiment, the Text Manipulation methodology includes parsing the input data object by locating the defined verbiages for every snippet. The readable data object created post OCR on the structured or unstructured input data object may contain all types of information like machine printed text, handwritten text, seals or stamps & signatures, logos etc. Solution has logic to identify the seal/logo or specific machine printed text. Based on the defined metadata - the verbiage associated with snippet is located to determine its exact location. With this location information the solution refers back the original image and renders the required snippet information (informative segment).
[0036] In order to retrieve one or more close results of the informative segments with respect to the customized search query the classification information with respect to

the data object needs to be configured. The processing engine (106) configures the data object classification information with respect to the data type of the data object or data object contents. By way of a specific example, the classification information for a document may include but is not limited to the image type like tiff or jpg. Moreover, the processing engine (106) may retrieve the results by searching through structured, semi-structured or unstructured data object.
[0037] After the customized search query is configured and processed, the processing engine (106) retrieves one or more close results of the informative segments with respect to the customized search query since the query is searched in more than one data store.
[0038] As shown in figure 1, the processing engine (106) further comprises of the validation module (110) which confirms an accuracy of the informative segment thus recognized by referring to the data object in the original format. The informative segment thus recognized and extracted in as closest result is compared with the data object in the original format to check the accuracy of the segment. This step is completely automatic and connection to original data store (once the results are obtained) is performed by the integration module. The integration module (114) is also capable of recognizing the data due to its advanced configuration such that it provides a connection with the relevant data stores only rather than connecting to all the data stores present therein.
[0039] Still referring to figure 1. the intelligent system (100) further comprises of the output generation module (112) to retrieve one or more validated close segments of object results with respect to the user's search query (as shown in step 210 of figure 2). This output generation module may further include a display device or some similar hardware entity.
[0040] The list of close results of validated informative segment may include one or more results from which the user has to select the most correct or accurate result. The intelligent system (100) also provides the user an access to view the mapping/validation of all the close results with those of the data object in original

format so that it becomes easy for the user to select the correct result with correct informative segment out of the list of results. [0041] In accordance with an alternative embodiment, if the results obtained are not relevant, the user may also customize the search query (which is re-processed) by changing the input parameters for which the intelligent system (100) will also check the accuracy/relevance of list of new input parameters thus created by user and accordingly accepts the suggestive measures thus provided by the user.
BEST MODE/EXAMPLE FOR WORKING OF THE INVENTION
[0042] The intelligent system and intelligent method illustrated to recognize and extract an informative segment from an object data may be illustrated by working example stated in the following paragraph; the process is not restricted to the said example only:
[0043] Let us consider in a government agency (LIC office, passport office etc) user A's identity proof and a document containing his sign (manual) is stored in a system/data store of some branch at some other location. One of the office assistant before issuing an approval letter wants to crosscheck the details from the documents with those written in the form.
[0044] The intelligent system and intelligent method of the present invention provides to the user a way to define his search query in order to search for the information from a machine converted format of related document. The assistant then prepares a search query by entering certain inputs parameters like (form no, issued number, passport number etc) and enters his search via a user interface. Let the prepared query includes name - Ashok and Passport number 0056.
[0045] The processing engine of the intelligent system then maps the search query (name and passport number) with plurality of data objects thus stored in all of the integrated data stores in all the branches and certain results are obtained. Let 2 documents are

obtained with names Ashok and Ashoka and passport numbers 0056 and 005 respectively.
[0046] Further to sort out of these results and to select the accurate one, the validation module of the processing engine checks the accuracy of all the results by referring to data in original format (where the picture is also present) and rejects those which are not correct. Let the processing engine considers both the results as close results of validated informative segments.
[0047] Both these validated results are displayed to the assistant along with the mapped informative segment with that of the original document so that it gets easy for the assistant to ensure over the accuracy of the document if the results are more than one or even one. The assistant by referring to mapped keywords clearly identifies which is the correct result and thus selects the result which comprises of name Ashok and passport number 0056.

WE CLAIM:
1. An intelligent system facilitating recognition and extraction of an informative segment from an object data, the intelligent system comprising: a user interface configured to define a customized search query by receiving one or more input parameters from a user;
a data conversion module configured to convert original format of said object data into a machine readable format and store the machine readable format in a respective data store, in order to search and recognize the informative segment from said object data with respect to the customized search query of the user;
a processing engine by using a set of programmed instructions, configured to retrieve one or more close results of the informative segment with respect to the customized search query by searching in one or more data store, the processing engine further comprising;
a data mapping module configured to map the input parameters of the customized search query with the data stored in the respective database by using a text manipulation methodology and further parse through said data object in order to recognize the informative segment;
a validation module configured to check an accuracy of the informative segment thus recognized by referring to the object data in the original format; an output generation module configured to retrieve one or more close results of the validated informative segments of object data with respect to the customized search query.
2. The intelligent system of claim 1, wherein the object data further comprises of scanned images, machine printed text and a combination of machine printed text and hand written data.
3. The intelligent system of claim 1, wherein the data conversion module further implements an OCR (Optical Character Recognition) technique.
4. The intelligent system of claim I, wherein the informative segment further comprises of a text part of the object data.

5. The intelligent system of claim 1, wherein the original format of data further comprises of a structured data, an unstructured data, a semi-structured data and a combination thereof.
6. The intelligent system of claim 1 further comprises of an integration module to search within one or more data store.
7. An intelligent method facilitating recognition and extraction of an informative segment from an object data , the method comprising:
defining a customized search query by receiving one or more input parameters from a user in order to search and recognize the informative segment from the object data; converting original format of said object data into a machine readable format and store the machine readable format in a respective data store in order to search and recognize the informative segment from said object data with respect to the customized search query of the user;
processing said customized search query by using a set of programmed instructions to retrieve one or more close results of informative segment by searching in one or more data store, the processing further comprising;
mapping the input parameters of the search query with the data stored in the respective data store by using a text manipulation methodology and further parsing through said data object in order to recognize and extract the informative segment; checking an accuracy of the informative segment thus recognized by referring to the data object in the original format;
retrieving one or more close results of validated informative segments of object data with respect to the user's search query.
8. The intelligent method of claim 7, wherein the object data further comprises of scanned images, machine printed text and a combination of machine printed text and hand written data.
9. The intelligent method of claim 7, wherein the informative segment further comprises of a text part of the object data.

10. The intelligent method of claim 7, wherein the original format of data further comprises of a structured data, an unstructured data, a semi-structured data or a combination thereof.