FORM 2
THE PATENT ACT, 1970
(39 OF 1970)
COMPLETE SPECIFICATION
(See Section 10 and rule 13)
SYSTEM AND METHOD FOR IMPROVED JOB SEARCHING"
Propellum Infotech Pvt. Ltd
A company incorporated as per the laws of India
With their registered office at: 301, B Wing, Third Floor, Time Square Building, Andheri-Kurla
Road, Andheri East Mumbai -400059, Maharashtra, India
The following specification particularly describes the invention and the matter in which it is to be performed.

SYSTEM AND METHOD FOR IMPROVED JOB SEARCHING
Field of the Invention
The invention relates to electronic access to job information by crawling URLs. More particularly, the invention relates to a method and system to identify relevant URLS specific to the job required by the user and after identifying those specific URLS the system enables to crawl them to fetch specific jobs.
Background of the Invention
Various job search vehicles are available to a job searcher or recruiter today on the Internet, e.g. the World Wide Web (web). However, the available resources that can be scrutinized are somewhat limited. There are job crawling engines for crawling over various websites to provide details of jobs. However the job crawling engines crawl over all the possible job websites and dump high volume of data for further scrutinization to fetch relevant jobs required by the user.
Accordingly there is a need for a system that more completely searches and organizes information that can be obtained from as many sources as possible via the Internet and present search results in an effective manner to a job searcher in response to a query and for assisting job seekers in their job search efforts by identifying, extracting, capturing, and leveraging expertise and knowledge.
2

Object of the Present Invention
The object of the present invention is method of managing user's preferences and provide enhanced access via internet to the relevant job data.
A further object of the present invention is to provide a more thorough and precise searching of details of jobs required by user
Detailed Description of the Invention
Present invention relates to electronic access to job information by crawling .URLs. More particularly, the invention relates to a method and system to identify relevant URLS specific to the job required by the user and after identifying those specific URLS the system enables to crawl them to fetch specific jobs.
The present invention provides a portal system for searching and managing job profiles as an interconnection of at least one configuration and namable business process element wherein said system comprising of a host computer, at least one database component associated with the host computer system, and a communication interface for accessing the host computer system from a plurality of remote input/ output devices to search and-manage the business process element,
The system for searching and managing job profiles further comprising a user job profile search input creation system to provide user inputs for analyzing various URL whether it contains relevant job description as per the user input or not. Thus, in general present invention consist of essential embodiments wherein user filling up the mobile based/ web based forms by the user while accessing through input output devices like mobile, iPhone, iPad, Android device, Windows device, Blackberry and any web based devices.

The present invention provides ease to user for searching relevant job data wherein the user enters specific job requirement on the web enabled device or devices; further, said user enters the URL for analyzation (whether it contains specific job or not) on to the web enabled device / devices wherein said device is connected to the portal carries the request for analyzation before crawling on a particular website. Therefore, the inputs provided by the user on web based form is transformed and analysed by the rule creation engine in the manner as described herein.
In one of the embodiment of the present invention the queue decider decides as to whether a particular URL contains specific required job content or not for being crawled based on the type of trainings. Thus the system analyze the words of the URLs against the system trained patterns for generating preliminary score. Further it evaluates based on the knowledge base patterns to generate a total score, based on that the system decides whether the URL needs to be crawled or not.
Thus the present invention provides three types of training which perform evaluation, therefore the Evaluation by Type 1 and Type 2 training returns the score after evaluating the URL by type 1 and type 2 training whereby a positive output from this evaluation means the URL would be sent for further evaluation by Type 3 training or else the URL is not considered and its deemed to be ignored thereby.
In an embodiment of the present invention wherein Type 1 training is based on string tokenization, wherein each training is given either PRO [Positive] or CON [Negative] score, the training is stored in the form of small tokens. Thus, the evaluation is made by comparing tokens with the input and final score is computed according to the positive and negative score given for the training.

For illustration purpose, if the following URL is being evaluated against type 1 training, https://www.abccopmanv/aboutus/careers/iobs/ioblisting therefore, from this URL only the part after the domain (i.e carrer/jobs/joblisting) is divided into tokens and evaluated against the trained tokens. Therefore the inputs given by the user is transformed into token.
EXAMPLE 1
Thus, assuming "careers and jobs" are trained as pro training and "about" is trained as con. These keywords are stored in the form of 3 character tokens in left to right order. So In this way following are the tokens created for the trained words:

Trained word Set of tokens stored
Careers Car,are,ree,eer,ers
Jobs Job, obs
About Abo, bou, out
EXAMPLE 2
The URL part to be evaluated is further matched against said tokens during evaluation and the total score is generated. Thus, for example if 1 pro token carries a total score of 1, and -1 for each con token the scoring of above URL will take place as shown in following table

Token found Score
Abo -1
Bou -1
Out -1
Car 1
Are 1
Ree 1
Eer 1

job 1
obs 1
Job (found 2nd time in job-listing) 1
Total Score 4
In an embodiment of the present invention wherein Type 2 training is based on full word, each training is given either PRO or CON score, full words are found in the URL and the final score is computed. Thereby the final score is computed after combining the evaluation result by both the training.
EXAMPLE 3
Considering the same URL for type 2 training, and assuming that the word "joblisting" is trained as pro with weightage 3 and aboutus is trained as con with weightage -1. Thus, the Type 2 evaluation will take place as shown in following table

Word found Score
Aboutus -1
Joblisting 3
Total Score 2
As indicated, unlike type 1, there is no tokenisation in type 2. Further, type 1 and type 2 evaluation scores are combined and checked against a fixed threshold value. If the score is higher than the fixed threshold value the URL is sent for evaluation against type 3 training otherwise it is considered that the URL is not a job page nor does it lead to one. Hence it's not sent for crawling.
In an embodiment of the present invention wherein evaluation by type 3 training is disclosed
wherein, said evaluation returns score after evaluating the URL by type 3 training. In type 3 training it recognizes the suffix and prefix to ascertain whether it is a job page or not for

example, like job/id/job/career. Thus a positive output from this evaluation infers that the URL would be sent to higher priority else the URL will be sent to Medium / Normal priority queue.
EXAMPLE 4
If said URL is sent for type 3 evaluation wherein the PRO keywords are "jobid, jobdetails, viewjob" and as these keywords are not in the URL it will not return a positive score. Hence, the URL will be considered as not a job page and will not be sent to higher priority. However, when the following URL found as a link on this page is sent for type 3 evaluation the scoring will be computed as below: URL: https://www.abccopmanv/aboutus/careers/iobs/jobid=12543
In this case the pro word "Jobid" is found and the page is considered as a job page and sent to higher priority (i.e data extraction). In this training a single CON word or absence of any PRO word is considered as a negative output and the URL is not sent for higher priority/There is no weightage consideration in this case. That is if a URL has 4 PRO words but 1 CON word the outcome is still considered to be negative.
Thus, the system analyses the URL through the method of trainings and based on the score, the system suggests to crawl that website or not. The URL, entered by the requester is analysed through the system using said method of trainings and then the system analyses the sores with the tolerance level scores and accordingly transmits the results to the user on its web enabled device, the system indicates the requester by an alert that the URL contains relevant job specifications, and if the requester clicks 'go ahead' then the system crawls that website and fetches relevant job details and provide the same on the portal of the system and is transferred to

the user on the web enabled device. Throughout this disclosure the web enabled devices are any of the devices such as mobile, iPhone, iPad, Android device, Windows device, Blackberry and the like.
Although functional components, modules, software elements, hardware elements, and features and functions described herein may be depicted or described as being fixed in software or hardware or otherwise, it will be recognized by persons of skill in the art that the features and functions described herein may be implemented in various software, hardware and/or firmware combinations and that the functions described herein may be distributed into various components or subcomponents on the network and are not fixed to any one particular component as described herein. Thus the databases described may be separated, unified, federated, or otherwise structured to best suit the preferences of the implementer of the features and functions described herein. Also, the functions described herein as preferably being performed manually maybe performed manually or maybe divided into subtasks which may be automated and ultimately performed by intelligent subsystems which mimic human operator interaction such as artificial intelligence systems which may be trained by human operations and ultimately function autonomously. Further features, functions, and technical specifications are found in the attached descriptions further below as well as claims contained therein.
If the user has entered inputs that he needs to know about Zycus company related jobs pertaining to product designing department then the above training patterns and knowledge patterns and rule creation engine will hit Zycus related original domain, however many a times the original domain directs to some other domain like Pebstone company or Propellum company

Whenever the URL link directs the crawler to some other domain other than the original domain, there is a possibility that other domains may contain jobs related to the original domain or may not contain any related jobs. In the following circumstances the method to crawl in a strategic way by having various test criterias.
Start
- Input is the Home URL [ From which the main crawling starts ] , URL which belongs to other domain, Company Data and Crawling Configurations
- Cross domain crawling is done only when there are no jobs found in the home domain.
Threshold Test while Crawling
- While crawling, if the number of consecutive pages with no jobs found increases 100 then the crawling is proposed to be terminated for the domain.
- Another threshold is to check whether the number of jobs found exceeds the number of expected jobs derived from company's strength with some ratio specified. If the threshold is crossed during the cross domain crawling, the crawling for the domain terminates. Else the crawling continues

We claim:
1. A system and method for searching and managing job profiles as an interconnection of at
least one configurable and namable business process element; said system comprises:
a host computer system;
at least one database component associated with the host computer system;
a communication interface for accessing the host computer system from a plurality of remote input/ output devices to search and-manage the business process element, further comprising:
a user job profile search input creation system to provide user inputs as source for analyzing various URLs connected to the source, whether it contains relevant job description as per the user input or not.
2. A system and method as claimed in claim 1, wherein the user enters specific job
requirement on the web enabled device or devices; further, said user enters the URL as
source for analyzation of all other URLs connected to the source (whether it contains
specific job or not) on to the web enabled device / devices wherein said device is
connected to the portal carries the request for analyzation before crawling on a particular

website.
3. A system and method as claimed in claim 1 wherein, the inputs as source provided by

the user on web based form is transformed and analysed by the rule creation engine

wherein, the queue decider decides as to whether a particular URL contains specific
required job content or not for being crawled based on the type of trainings.

While the system and method have been described in terms of what are presently considered to be the most practical and preferred embodiments, it is to be understood that the disclosure need not be limited to the disclosed embodiments. It is intended to cover various modifications and similar arrangements included within the spirit and scope of the claims, the scope of which should be accorded the broadest interpretation so as to encompass all such modifications and similar structures. The present disclosure includes any and all embodiments of the following claims.

4. A system and method as claimed in claim 1 wherein, the system analyze the words of all the URLs connected to the source against the system trained patterns for generating preliminary score, further it evaluates based on the knowledge base patterns to generate a total score, based on that the system decides whether each URL connected to the source needs to be crawled or not.