Printer Friendly

EXTRACTING PARAMETERS FROM E-FEEDBACK TEXT DATA BASED ON KEYWORD BASED PARSING IN ELEARNING ENVIRONMENT.

Byline: Muhammad Farhan Ramash Zahra Muhammad Munwar Iqbal and Muhammad Aslam

ABSTRACT: The parser utilized is a variant of a Shift-Reduce parser. It really contains three diverse sensible movements that the parser utilization to get to the last parse which is a semantic understanding (in first-request rationale) of a common dialect articulation. This left to right parsing makes the methodology moderately instinctive for human Resume Parser may mean extraction of abilities capabilities and knowledge of a trail of occupation seekers from their resumes transferred in different employment entrances for capacity dissection and assessment by HR Managers. A Parser is a machine program that tries to recover the components of competitor's capability expertise and encounter alongside applicant's names and location from the content of resumes composed in different configurations and archives them at HR stations to dissect assess and shortlist seekers. Resume Parser has enormously soothed the

HR Managers from undertaking the difficult activity of recruitment which until its coming took a few weeks and months to lead the activity effectively. Searching text data based on keywords is a better technique in electronic learning (eLearning) environment.

Keywords: Keyword based parsing parsing search E-Learning electronic feedback (e-Feedback)

INTRODUCTION

Over the past few years internet search technology has progressed far beyond the traditional bag-of-words retrieval framework. In particular for some web search queries parsing-based Semantics is employed to parse" the query using some rules or Grammar and some underlying database of facts and this parse influence the search output. As a simple example for the query Seattle to Redmond typical search kingies use rules to identify that the query matches location to locations since Seattle and Redmond is both present in a location table. This parse once identified may then be consumed by a map application which displays a map with directions alongside the search results.

There are several advantages to the parsing-based approach: (1)The parse of a query provides clues to possible user intentions behind the query. For instance a query business customer support is presumably looking for customer support contact information for a business. A better understanding of the user intentions allows the search system to customize the output to a query e.g. Show a map of driving directions or display contact information. (2) A parsing-based approach allows the query to contain no content terms (e.g. from to) combine data and schema terms (e.g. best buy phone) and relate content both inside and outside of join paths" (e.g. godfather Brando " matching a single row in a movie-actor relation versus New York to Scottsdale " matching multiple rows in a location relation);

Such queries are hard to support in keyword search over databases that interprets all search terms as content arising from some row in the database. Thus parsing-based semantics allows internet search kingies to Override the bagof-words retrieval framework and provide internet users with better or enhanced search results. In fact prior work has shown that search results where such enhanced information is provided have an order of magnitude higher user satisfaction rate than those without to the best of our knowledge however the implementations of these semantics are ad-hoc and there has been

No formal study of the efficiency of the parsing-based semantics for internet keyword search.

Parsing keyword search queries is equally relevant in enterprise keyword search over database. Previous work at IBM on the Avatar project for enterprise search used heuristics to parse keyword search queries to a grammar. More recently Fagin provided a formal basis for the problem of answering queries using a (particular instance of) parsing-based semantics and showed that the problem admits a polynomial time (in input And output size) solution. While this result contains key insights into the theoretical complexity of the problem Fagin et al. did not propose efficient algorithms or indexes for the problem. For efficient keyword search indexing techniques that avoid repeated scans of the entire input database to answer a query are critical [19]. The central goal of this paper is to develop efficient indexing and query answering techniques to support parsing-based keyword search.

Our techniques are general; in particular we study how the following dimensions impact the efficiency of our approach as shown in figure 1.

Resume analyzer performs many estimations inside few seconds and records Candidates' information of work experience instruction and abilities into industry standard HR-XML format [1].

There are by and large three sorts of Parsers:

Keyword based parsers

Grammar based parser

Statistical parser

Pivotal word based parsing contrast and different sorts.

Keyword Search for XML Databases

The Xk look schema inputs a rundown of crucial words and outfits a relative payback of Smallest Lowest Common Ancestor center points i.e. the rundown of centers that are establishments of trees that hold the catchphrases and hold no center that is in like manner the base of a tree that holds the watchwords. For each vital word the schema underpins a rundown of centers that hold the catchphrase as a tree sorted by the id's of the center points [17]. The key property of SLCA interest is that given two keywords k1 and k2 and a center point v that holds vital word k1 one necessity not analyze the whole center point rundown of catchphrase k2 to discover potential results. Rather one simply needs to find the left and right match of v in the rundown of k2 where the left (right) match is the center point with the best (least) id that is more little (additionally astonishing) than identical to the id of v. The property totals up to more than two catchphrases and prompts the Indexed Lookup Eager count

whose essential memory unusualness is o (|s1|kd log|s|) where d is the most great significance of the tree k is the measure of watchwords in the request and s1|(|s|) is the base (most awesome) size of catchphrase records s1 through Sk as specified in figure 2 [67].

Keyword Queries over Web Knowledge Bases

Pivotal word look over XML reports with the point of distinguishing the most genuine substance components that hold all the information essential words alongside a minimal associated tree to portray how each one effect matches a given catchphrase question [16]. To gain all the more constraining impacts of vital word questions we propose the plans of Valuable LCA and Compact VLCA to accurately and profitably address XML watchword request. In light of the two thoughts we propose the preservationist partnered trees made Cvlcas as the answers of critical word request. Likewise we show a change technique for enlivening the handling of Cvlcas and devise a gainful stack-based computation to recognize the genuine preservationist copartnered trees.

We have actualized our system and the far reaching test effects indicated that our technique accomplished high effectiveness and adequacy on both manufactured and genuine datasets [9].

Assessing the dispersion in specifically would oblige a lot of marked preparing cases. The space of conceivable magic word inquiries mapping to conceivable KB questions is substantial and evaluating such a mapping straightforwardly is not achievable when preparing information is constrained. Due to the manual needed in making marked preparing information we have to plan an approach that can amplify the utility of a little gathering of preparing illustrations.

1. Essential word Query Annotation Queries are ist commented with the semantic develops from an information representation dialect (i.e. element sort characteristic esteem connection). We utilize grammatical form tags as characteristics that prescribe reasonable to assume semantic develops for each one question term. The mapping from grammatical form tags to semantic builds is gained from an expounded inquiry log.

2. Essential word Query Structuring Annotated inquiries are organized by processing the most likely organized question layouts given the annotations as a semantic outline of the inquiry substance. This relationship between annotations and inquiry structures is gained from an expounded question log. Taking in a mapping straightforwardly from essential words to organized inquiries might oblige a lot of preparing illustrations. By taking in the mapping from semantic outlines to inquiry layouts we exploit the repetition in the preparation information initiated by numerous inquiries offering the same synopses and formats.

3. Learning Base mapping semantically commented Essential word inquiries could be consolidated with an organized Question layout to structure an organized

TABLE 1: Key word based parsing compare with other types

###Keyword based parser###Grammar based parser###Statistical parser

###Catchphrase###Linguistic use based###Measurable Parsers endeavor to

###based parsers are the###parsers are guided by syntactic concentrate numerical figures from the

###least difficult and the###standards.###content of resume.

###minimum precise.###The linguistic use principles###Measurable parsers can accomplish a

###They attempt to###help the parser comprehend the high rate of exactness in information

###recognize###words###or###significance of each sentence extraction.

###expressions from given###and consequently separate###Measurable###parsers###can

###catchphrases.###significant data.###notwithstanding accomplish high correctness

###These parsers fall Linguistic use based parsers on information on which they are prepared

###lacking when the terms###have the ability to get a higher yet this is not generally extremely handy since

###of aptitude capability or###rate of exactness.###this information is by definition old

###experience held in the###information that won't be seen once more.

###resume are past given

representation of the essential word inquiry known as an organized catch phrase question. We stretch out an existing methodology to guide organized catchphrase questions into idea seek inquiries. 4. Information Base Query Evaluation Concept look Inquiries are then executed over the learning base to _nd elements and qualities depicted by the question. This Methodology performs question time derivation abusing the Semantics encoded in the information base to figure Inquiry responses utilizing a custom learning base motor (Examination of the learning base motor is past the extent of this paper however any database framework that can actualize the semantics. and elements and qualities depicted by the inquiry [4]. This methodology performs question time induction abusing the Semantics encoded in the information base to process inquiry responses utilizing a custom learning base motor.

1. Grammar: A central component of query parsing is the grammar that specifies how queries are parsed. We consider a rich class of grammars that can be loosely characterized as a regular expression over database concepts (columns) and special keywords (e.g. to from).

2. Matching: To parse a query we need to determine if a particular phrase in a query matches a value in the database. Given that keyword queries are typically short and incomplete requiring exact Matches would not be robust. We therefore focus on set containment as a matching function to determine matches which is widely used in traditional keyword search as well. We also consider and present results for other well-known matching functions such as approximate equality and substring containment.

3. Scoring: Again to be able to handle short and imprecise queries we work with a relaxed definition of a parse that can skip over and cover" only a subset of query terms. With this relaxed definition often a query can be parsed in a multitude of ways. In this setting it is important to be able to capture preferences for one parse over another. A natural preference is to prefer parses that cover more queries Terms to parses that covers less. We consider a general class of scoring functions to express such preferences and this class includes well-known sequential models such as CRFs and HMMs [2].

SURVEY OF SEMANTIC SEARCH APPROACHES

We are seeing a huge advancement in the measure of information that is instantly available readily available with the impact of the World Wide Web or the web. Web removes the physical limits that were once associated with learning disseminating making an individual's physical territory all things considered insignificant [20]. Ebb and flow examination and insightful articles that were once accessible to just a chosen few can now be made accessible for everybody on the web. Once made accessible on the web any scholar in any some piece of the world with access to the web can gain access to these articles to improve their learning. The gigantic measure of data accessible on the web accompanies its set of tests. One of these tests is the test of data recovery [8]. The incomprehensible measure of data accessible on the web must be utilized successfully with effective data recovery. Given a client seek as an inquiry data recovery tries to return connections to a set of articles that are likely of service to

fulfill the client need for data. An assortment of the inquiry calculations resort to straightforward catchphrase based quests. The client question is part into a situated of catchphrases and an animal energy inquiry of the article database is made to get a set of matching effects. The amount of comes about that are returned can regularly shift a substantial sum relying upon the level of specialization of the client specified decisive words. To enhance client experience web indexes resort to a mixture of positioning procedures for example Page rank Citation Indexing to present the query items. The World Wide Web as it exists today could be viewed as an extensive semi organized database [12] [13]. The worldwide benchmarks figure has as of late assumed the undertaking of a community development towards a Semantic Web. The benchmarks advertise the utilization of institutionalized information groups for data on the web.

The principle point of the semantic web is to permit clients to discover and portion data all the more effortlessly. As a parallel semantic pursuit systems attempt to comprehend the real client goal of an inquiry so they can return better indexed lists to the client [14] [15].

It likewise records four separate methodologies to perform a semantic inquiry. The primary methodology utilizes connection data to translate the aim of the client inquiry. As a sample think about the expression bark. Contingent upon the connection of the client question it can allude to two separate words. The point when utilized as a part of the connection of 'canine bark' it implies the short boisterous shout of a pooch. The point when utilized as a part of the setting of trees it alludes to the intense outer surface spread of a root or stem. Relevant investigation endeavors to enhance the list items by making utilization of the connection data of an inquiry. The second approach utilization thinking. An inquiry framework that makes utilization of thinking to enhance it comes about thinks about connections around articles and how to surmise new connections from existing connections. It utilizes this information to give better list items.

The third approach utilizes characteristic dialect handling. Internet searchers utilizing this methodology at-entice to parse a client inquiry and reason data from the same to recognize things for example individuals and spots. Such web indexes catch things like what was the subject the item and connections around the words. The point when the client inputs an inquiry it endeavors to match this semantic data with the semantic data of the web article. A case of this might be inquiries like 'who prevailed over Ravana' vs `ravana vanquished who' where the plan of the two questions is completely distinctive and a straightforward pivotal word based pursuit won't be completely exact as both questions will probably furnish a proportional payback set of effects. In any case a regular dialect transforming framework might have the ability to translate the goal of such inquiries precisely to touch base at better query items. The last approach makes utilization of metaphysics for space particular quests.

Philosophy formally speaks to information as a set of ideas in an area. Metaphysics might be utilized to model the space the sort of questions that exist and their properties or relations. As an illustration a metaphysics that is utilized to model vehicles realizes that an auto is a sort of a vehicle and it can utilize this connection to expand or sum up the search [3].

JOINING KEYWORD BASED SEARCH WITH SEMANTIC SEARCH

As a while ago specified semantic ventures ordinarily experience the ill effects of adaptability issues. Decisive word built inquiries with respect to the next hand scale well as the database size develops. It is additionally conceivable to parallelize decisive word based quests to get further upgrades in execution in a straight send and simple manner utilizing C++ dialect expansions for example Intel Cilk Plus It might be decent in the event that we can consolidate the two ventures so we can exploit adaptability of decisive word based pursuits alongside the likelihood of getting more faultless outcomes from the semantic inquiry approach. In this segment we introduce the subtle elements of our work to execute a mix of the two pursuits. With a specific end goal to exhibit the suitability of our methodology we think about a database of productions. The database might be recognized as having insights about every distribution alongside a connection to the real production.

Every production has a title and is clarified with a few bits of semantic data for example if it was distributed in a meeting or a diary the year it was distributed in the creators of the distribution and the guardian collection of the diary or the gathering for example IEEE or ACM. An entrance in the database might be demonstrated pictorially as takes after.

The inquiry calculation begins by doing a preparing of the client question to endeavor to comprehend the client goal. Think about the client inquiry: IEEE diary production on semantic hunt by Joe Someone in 2009. To keep our execution straightforward we search for specific essential words to figure the client plan. Certain words for example IEEE ACM are dealt with as unique and are acknowledged as defining the guardian association in.

The pseudo code for the calculation looks as appeared [3]. Combination_keyword_semantic_search(user_quer y_string)

parse_output = pre_parse_query(user_query_string) for each one article in the database assuming that parse_output-greater than search_keywords are found in article-greater than title add article to candidate_set end if

end for

each one article in the candidate_set num_semantic_matches = 0 assuming that article-greater than author matches parse_output-greater than author increase num_semantic_matches end if

assuming that article-greater than type matches parse_output-greater than type increase num_semantic_matches end if

assuming that article-greater than publication_year matches parse_output-greater than publication_year increase num_semantic_matches end if

assuming that article-greater than parent_organization matches parse_output-greater than parent_organization increase num_semantic_matches end if

assuming that num_semantic_matches == 0 proceed else

add article to a vector comparing to num_semantic_matches end if

end for if showing all_match list items presentation articles in vector comparing to max_allowed_semantic_pieces else

for( num_match = max_allowed_semantic_pieces;num_match-;num_match greater than = 1)presentation articles in vector comparing to num_match end for

end if

End

Google applies two essential elements for judging the criticalness and importance of any site page before positioning them. These elements are Page Rank (for measuring prevalence by investigating the backlinks) and pertinence (by examining the utilization of watchwords or inquiry question terms utilized as a part of the site page). However this manifestation of positioning reports don't assist to discover those pages which may be pertinent to searchers purpose as the prevalence variable may lessen the rankings of semantically significant records. This is the reason that Google utilizes semantics to recognize and prioritize the rankings of pages having semantically pertinent substance as opposed to just numbering the watchwords and back connections for dissecting any site page.

Inquiry Processing In a Semantic Environment

The figure underneath portrays the steps included in the preparing of the inquiry by Google. The hunt question gained by Google is parsed (utilizing a parser) to distinguish one or more parts (first and second pursuit terms). In this procedure equivalent words or other displacement terms gets distinguished. These equivalent words are known as competitor equivalent words and they further get broken down and handled as qualified equivalent words. At that point a relationship motor is utilized to distinguish the relationship between the parts based upon their separate areas. Here an area essentially implies a brought together classification of comparable words. To begin with pursuit term gets recognized by the first area which is a semantic class having an accumulation of predefined substances. Thus the second term gets recognized by a second area likewise holding a database of comparative elements.

This helps Google to relate the terms to the closest matching characters (One vital poin to note here is that Google will just discover and relate words in the question with those generally exhibit in its database which is the Knowledge Graph henceforth a few inquiries despite the fact that semantically comparable may not appear). A separate hunt gets led by a question motor utilizing area matching relationship (don't get confounded with the statement space with area name here area implies class) and last comes about gets showed after a semantic inquiry is distinguished (the inquiry motor may pluralize or reword the inquiry if needed). Thus in straightforward words a complex inquiry entered by the client is broken down and rearranged including a few methodologies into semantic question. From there on applicable website pages are recognized and showed in figure 3 as a last set of outcomes.

Keyword search over structured data such as relational Databases are an undeniably vital ability exploiting a mixture of DB and IR strategies. While these activities concentrate on watchword based inquiry preparing in an unified database the infolding sending of P2p systems and administration arranged architectures has made it just as paramount to amplify such catchphrase based pursuit proficiencies to appropriated databases. Similar to dispersed IR

Frameworks essential word based database choice is a discriminating step towards spotting helpful databases for noting a catchphrase question and on which existing unified watchword seek strategies could be straightforwardly connected.

For powerful determination of valuable information sources in distributed.ir frameworks a typical methodology is to compress archive accumulations with an arrangement of catchphrases connected with some intra-gathering (e.g. recurrence) or between accumulation weightings. Information sources are then positioned by contrasting the watchword questions and their rundowns which might be put away either at the information sources or the questioning customers Summarizing a social database with the basic magic word rundown technique as in IR frameworks may be be that as it may deficient for two reasons. To begin with social tables in a database are commonly standardized. In this manner the magic word recurrence detail which are utilized as a part of most IR-based outlines for printed records can't generally measure the significance of decisive words in a social database. Consider the situation where a pivotal word seems just once and it is in a tuple that is referenced by numerous different tuples.

Such an essential word is liable to be imperative since it is identified with numerous different catchphrases in the associated tuples. Second the results from a social database as for a catchphrase question must consider the amount of join operations that must be carried out in place for all the essential words to show up in the result (regularly spoke to as an assessment tree). This must be acquired if the relationship between decisive words in the social database is by one means or another caught in the synopsis. For delineation let us take a gander at the two case databases Db1 and Db2 indicated in Figure 1 in which the arrowed lines drawn between tuples show their associations focused around remote key references. Assume we are given a magic word question Q = f sight and sound; database; VLdbg [10] [11]. We can watch that Db1 has a decent come about to Q.

We make the accompanying commitments in this paper. We propose to take a gander at the issue of organized information sources choice for decisive word based inquiries. To the best of our learning this is the rest endeavor to address this issue.

We propose a system for abridging the relationship between pivotal words in a social database. The strategy for creating the database synopsis is possible by issuing SQL proclamations and along these lines could be performed straightforwardly on the DBMS without alteration to the database motor.

We done measurements for electively positioning source databases given an essential word question as indicated by the catchphrase relationship outline joining tuple t1 with t3. Unexpectedly Db2 can't give applicable results to Q there are no trees of associated tuples holding all the question magic words. Anyway on the off chance that we assess the two databases for Q focused around the decisive word recurrence style synopses (signified as KFsynopsis in this paper and KF-summary(db1) = { multimedia:1 database:2 Vldb:1} and KF-summary(db2) = { multimedia:3 database:3 Vldb:1 }) Db2 will be chosen over Db1. Accordingly we can watch that the convenience of a social database in noting a magic word inquiry is not just chosen by whether it has all the question essential words however all the more significantly it relies on upon whether the inquiry catchphrases could be joined genuinely in the database. In this paper we dene pivotal word relationship for speaking to such associations between magic words in a social database and take a gander at how condensing catchphrase connections can help us to eagerly select important organized sources in a disseminated setting. This work is a piece of our Best Peer venture for supporting P2p-based information offering administrations. Best Peer is a P2p stage that might be invoked to help either organized or unstructured overlays and it gives a set of apparatuses to building information offering requisitions [18].

Current methodologies for preparing watchword based questions over XML information think about inquiries that point out stipulations on the printed substance or on both the literary and the marks of every XML component. Therefore we expect a straightforward magic word based inquiry dialect with grammar acquired A catchphrase based question q over a XML archive stream is a rundown of inquiry terms (additionally meant hunt terms) [21]. Each one question term is of the structure: l::k l:: :: k or k where l is a component mark and k a catchphrase. A hub n inside an archive d fulfills a question term of the structure: l::k if n's name is equivalent to l and the text based substance of n holds the watchword k; l:: if n's mark is equivalent to l::k if the text based substance of n holds the pivotal word k; k if n's name is equivalent to k or the literary substance of n holds the catchphrase k.

We additionally expect that the magic word based inquiries are assessed against records utilizing the well-known SLCA semantics1.

To delineate our question dialect think about the accompanying inquiry determinations:

sa: New books composed by Doorman

sb: New books having "Doorman" in their title

sc: New books in Mp3 arrange by distributer named "Mp3"

Table 1 presents conceivable catchphrase based question details that satisfy these inquiry specifications2. These watchword inquiries were deliberately assembled to show our calculations and are focused to streams. Holding book discharges. Additionally they represent diverse situations with respect to the client's information about the XML names. 1actually any viable LCA-based semantics could likewise be utilized.

2. Traits could be effectively took care of with no huge overhead. Strategy Multi Inquiry

Information: A stream of XML reports D A set of inquiries Q (from clients' profiles) are handled against a stream D of records. Upon its landing each one record dj in D is independently transformed (Lines 3 to 6). The outcomes found inside this record are gathered and returned. This is refined in the meantime for all inquiries qi and outcomes are separately gathered in every ri. A consequence incorporates the full way and the Dewey Code of the SLCA coming about hub of dj if any.

Each one record dj in D is prepared by a SAX parser which produces five sorts of occasion for an archive: begin Report ( ) begin Element (tag) characters (text) end Element (tag) and end Record ( ). Our calculations then work by method for SAX Callback Capacities for those occasions. The parser is called to movement in Line 5 of the method.

Parsing stack. As the SAX parser navigates the archive in an in-place form each one went to hub is connected with an entrance in a stack S which we call the Parsing Stack. Every passage is popped for the stack when it's relating hub and all its relative have been gone by. To backing the recursive preparing of a record every passage in the parsing stack handles the accompanying data:

(1) The mark of the component relating to the passage;

(2) A bitmap called Could BE SLCA which holds one bit for each one question qi being assessed

(3) A set utilized questions holding the Ids of the inquiries whose terms incorporate pivotal words introduce in the component (or its relatives) either as names or as in qualities;

(4) Which decisive words from these questions have happened in the relating record hub and its relatives Recognize that without misfortune of simplification we expect that names are exceptional for every sort of component in each one record. For example in one single report the name Writer is constantly used to speak to a book writer and not a paper author [5].

Question File. Throughout the traversal of a report it is important to search for catchphrases that happen in content components or names. As we hope to process an extensive number of inquiries our calculations depend on question lists so as to abstain from finding each one question independently. The indexing structures we utilize are adjustments of altered records. Each one file entrance speaks to a question term and alludes to inquiries in which this term happens making a qualification between structural (mark) and non-structural (worth) question terms. Perceive that as in any rearranged rundown inquiry records are produced in advance from the set of inquiries postured by clients. Handling new inquiries obliges the inquiry lists to be remade.

The stack when it's relating hub and all its relatives have been gone by. To backing the recursive preparing of a record every passage in the parsing stack handles the accompanying data:(1) The mark of the component relating to the passage;

(2) A bitmap called Could BE SLCA which holds one bit for each one question qi being assessed

(3) A set utilized questions holding the Ids of the inquiries whose terms incorporate pivotal words introduce in the component (or its relatives) either as names or as in qualities;

(4) Which decisive words from these questions have happened in the relating record hub and its relatives Recognize that without misfortune of simplification we expect that names are exceptional for every sort of component in each one record. For example in one single report the name Writer is constantly used to speak to a book writer and not a paper author [5].

Question File. Throughout the traversal of a report it is important to search for catchphrases that happen in content.

REFERENCES[1].GACopyrightnACopyrightreux M. (2002). An example-based semantic parser for natural language.proceedings of EMCSR2002.

[2].Parameswaran A. Kaushik R. and Arasu A. (2012). Efficient Parsing-based Keyword Search over Databases. [3].Narender G. and Rao M. S. Improving Search Accuracy by Combination of Keyword based Search with Semantic Information Search.

[4].Yu B. Li G. Sollins K. and Tung A. K. (2007 June). Effective keyword-based selection of relational databases. In Proceedings of the 2007 ACM SIGMOD international conference on Management of data (pp. 139-150). ACM. [5]. Hummel F. C. da Silva A. S. Moro M. M. and Laender A. H. (2011 October). Multiple keyword-based queries over XML streams. In Proceedings of the 20th ACM international conference on Information and knowledge management (pp. 1577-1582). ACM.

[6].S. Agrawal S. Chaudhuri and G. Das. DBXplorer: A system for keyword-based search over relational databases. In ICDE2002

[7].Xu Y. and Papakonstantinou Y. (2005 June). Efficient keyword search for smallest LCAs in XML databases. In Proceedings of the 2005 ACM SIGMOD international conference on Management of data (pp. 527-538). ACM. [8].Pound J. Hudek A. K. Ilyas I. F. and Weddell G. (2012 October). Interpreting keyword queries over Web knowledge bases. In Proceedings of the 21st ACM international conference on Information and knowledge management (pp. 305-314). ACM.

[9].Li G. Feng J. Wang J. and Zhou L. (2007 November). Effective keyword search for valuable lcas over xml documents. In Proceedings of the sixteenth ACM conference on Conference on information and knowledge management (pp. 31-40). ACM.

[10]. Sattler K. U. Geist I. and Schallehn E. (2005). Conceptbased querying in mediator systems. The VLDB Journal 14(1) 97-111.

[11]. Ramesh A. Sudarshan S. Joshi P. and Gaonkar M. N. (2013). Keyword search on form results. The VLDB Journal " The International Journal on Very Large Data Bases 22(1) 99-123.

[12]. Chen Y. Wang W. Liu Z. and Lin X. (2009 June). Keyword search on structured and semi-structured data. In Proceedings of the 2009 ACM SIGMOD International Conference on Management of data (pp. 1005-1010). ACM

[13]. Termehchy A. and Winslett M. (2011). Using structural information in XML keyword search effectively. ACM Transactions on Database Systems (TODS) 36(1) 4.

[14]. Bao Zhifeng et al. "Effective xml keyword search with relevance oriented ranking." Data Engineering 2009. ICDE'09. IEEE 25th International Conference on. IEEE 2009.

[15]. Bao Z. Ling T. W. Chen B. and Lu J. (2009 March). Effective xml keyword search with relevance oriented ranking. In Data Engineering 2009. ICDE'09. IEEE 25th International Conference on (pp. 517-528). IEEE.

[16]. Zhou R. Liu C. and Li J. (2010 March). Fast ELCA computation for keyword queries on XML data. In Proceedings of the 13th International Conference on Extending Database Technology (pp. 549-560). ACM. [17]. N. Bruno N. Koudas and D. Srivastava. Holistic twig joins:optimal xml pattern matching. In SIGMOD pages 310321 2002.

[18]. S. Chen H. Li et al. Twig2stack: Bottom-up processing of generalized-tree-pattern queries over xml documents. In VLDB 2006

[19]. F. Liu C. Yu W. Meng and A. Chowdhury. Effective keyword search in relational databases. In SIGMOD pages 563574 2006.

[20]. Z. Liu and Y. Chen. Identifying meaningful return information for xml keyword search. In SIGMOD 2007.

[21]. J. Lu T. W. Ling C.-Y. Chan and T. Chen. From region encoding to extended dewey: On efficient processing of xml twig pattern matching. In VLDB pages 193204 2005.
COPYRIGHT 2014 Asianet-Pakistan
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2014 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Publication:Science International
Article Type:Report
Date:Sep 30, 2014
Words:5756
Previous Article:INVESTIGATION FOR SAFETY PERFORMANCE INDICATORS ON CONSTRUCTION PROJECTS.
Next Article:DOES STOCK MARKET EXERT FINANCIAL DISCIPLINE ON FIRMS EVIDENCE FROM INDIA.
Topics:

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters |