Extracting two words from the article in english

i have an english article which is to be classified into a particular category based on the keywords. There are lacks of keywords stored in database. What i have to do is to obtain the keywords from the article and match it from the database. if match is found then the article belongs to that particular category. This keyword matching i did for one word by using split(" "), but now i want to do for 2 words from an article.that is getting 2 words from the article which is repeated many times.then searching it in the db.(here 2 words will be considered as one keyword)
Now what i should do to get the two appropriate keywords from the article without taking a,am,the,is,when etc...(leaving many generic words).
Any help will be appreciated.

hi,
thanks for reply!
I know its a bad algorithm classify the article written in english only based on few words appearing in the article.
But what i want to do is first extract the words from the article leaving the generic words, then count the single word each.Then i am sorting the words based on count and taking the five words from the article which has highest count. Now i have the database where millions of keywords are stored. These keywords are refering to particle category
ie. if we consider a category as sports, then under this category i have many keywords stored in the database like cricket, football, worldcup,tennis... etc
Now if i search the appropriate word from the article it will be considered as keyword. then this will be searched in the database. if match is found then it means the article belongs to sports category.
Now problem is some times article can have two words which can be considered as one keyword and can be used to classify article in much better way.
The question is how to get such words from the article???
ex.. if Hero's Journey is combined word appearing many times in the article then this keyword can be used to classify the article much better than going for single word.
Can anybody help me in this regard.
Any help will be appreciated.

Similar Messages

  • Substring between two words from a clob

    How to extract a string between two words from a clob variable in PL/SQL?

    My requirement is to extract the soap envelope from a clob. In the below code l_response returned from the http request is a clob with the below format. I need to extract all the text that is between '<s:Envelope' and 's:Envelope>'. That means I need to get rid of the first 5 lines and the last line from the l_response. Can you please help me with the logic?
    --uuid:18cb22a2-11cc-43f4-bfea-c213da179d30+id=157
    Content-ID: <http://tempuri.org/0>
    Content-Transfer-Encoding: 8bit
    Content-Type: application/xop+xml;charset=utf-8;type="application/soap+xml"
    <s:Envelope xmlns:s="http://www.w3.org/2003/05/soap-envelope" xmlns:a="http://www.w3.org/2005/08/addressing"><s:Header><a:Action s:mustUnderstand="1">http://tempuri.org/IUpdateService/QueryUpdateLogRecordsResponse</a:Action><a:RelatesTo>urn:uuid:413f419c-f489-44ea-bd12-dff6f24a4d71</a:RelatesTo></s:Header><s:Body><QueryUpdateLogRecordsResponse xmlns="http://tempuri.org/"><QueryUpdateLogRecordsResult xmlns:i="http://www.w3.org/2001/XMLSchema-instance" xmlns:x="http://www.w3.org/2001/XMLSchema"><XObject.m_element i:type="x:string" xmlns="">&lt;QueryResult Count="2" NextStart="0" PreviousStart="0" Id="{AD62FD77-AFBE-4362-BBEF-695DA5D92640}"&gt;&lt;Columns Count="33"&gt;&lt;Column AttributeName="Id" 
    … 5 pages later…
    DateModified="2014-07-06 07:34:41.9129549-07:00" /&gt;&lt;/Records&gt;&lt;/QueryResult&gt;</XObject.m_element></QueryUpdateLogRecordsResult></QueryUpdateLogRecordsResponse></s:Body></s:Envelope>
    --uuid:18cb22a2-11cc-43f4-bfea-c213da179d30+id=157—
    DECLARE
       l_request             VARCHAR2 (4000);
       l_http_req            UTL_HTTP.req;
       l_http_resp           UTL_HTTP.resp;
       v_buffer              VARCHAR2 (32767);
       n_next_start_record   NUMBER (20) := 1;
       l_response            CLOB;
    BEGIN
       -- Call webservices. Works fine
       l_request :=
             '--uuid:e4c19840-745d-45b2-90ca-12d71be4cfd9+id=2'
          || CHR (13)
          || CHR (10)
          || 'Content-ID: <http://tempuri.org/0>'
          || CHR (13)
          || CHR (10)
          || 'Content-Transfer-Encoding: 8bit'
          || CHR (13)
          || CHR (10)
          || 'Content-Type: application/xop+xml;charset=utf-8;type="application/soap+xml"'
          || CHR (13)
          || CHR (10)
          || CHR (13)
          || CHR (10)
          || '<s:Envelope xmlns:s="http://www.w3.org/2003/05/soap-envelope" xmlns:a="http://www.w3.org/2005/08/addressing"><s:Header><a:Action s:mustUnderstand="1">http://tempuri.org/IUpdateService/QueryUpdateLogRecords</a:Action><a:MessageID>urn:uuid:413f419c-f489-44ea-bd12-dff6f24a4d71</a:MessageID><a:ReplyTo><a:Address>http://www.w3.org/2005/08/addressing/anonymous</a:Address></a:ReplyTo><a:To s:mustUnderstand="1">http://dexdb5/DexNETWebServices_4_0_0_4/UpdateService.svc</a:To></s:Header><s:Body><QueryUpdateLogRecords xmlns="http://tempuri.org/"><context xmlns:i="http://www.w3.org/2001/XMLSchema-instance" xmlns:x="http://www.w3.org/2001/XMLSchema"><XObject.m_element i:type="x:string" xmlns="">&lt;OnlineContext SystemId="'
          || g_system_id
          || '" SessionId="'
          || g_session_id
          || '" UserId="'
          || g_user_id
          || '" /&gt;</XObject.m_element></context><xQueryRequest xmlns:i="http://www.w3.org/2001/XMLSchema-instance" xmlns:x="http://www.w3.org/2001/XMLSchema"><XObject.m_element i:type="x:string" xmlns="">&lt;QueryRequest Start="'
          || p_next_start_record
          || '" Count="'
          || g_records_count
          || '" Distinct="0" OrderBy="" Condition="(oUpdateLog.DateCreated &amp;gt;= '''
          || p_last_load_time
          || ''')" ColumnInfo="1" /&gt;</XObject.m_element></xQueryRequest></QueryUpdateLogRecords></s:Body></s:Envelope>'
          || CHR (13)
          || CHR (10)
          || '--uuid:e4c19840-745d-45b2-90ca-12d71be4cfd9+id=2--';
       l_http_req :=
          UTL_HTTP.begin_request (g_query_updatelog_records_url, 'POST', 'HTTP/1.1');
       UTL_HTTP.set_header (l_http_req, 'MIME-Version', '1.0');
       UTL_HTTP.set_header (
          l_http_req,
          'Content-Type',
          'multipart/related; type="application/xop+xml";start="<http://tempuri.org/0>";boundary="uuid:e4c19840-745d-45b2-90ca-12d71be4cfd9+id=2";start-info="application/soap+xml"');
       UTL_HTTP.set_header (
          l_http_req,
          'VsDebuggerCausalityData',
          'uIDPo5F/qXRc4YJImqB6Ard30cQAAAAAAjIXinpIVUulXLJOsSG7yyv7Lf2yHgpHlIxvc6oeqaAACQAA');
       UTL_HTTP.set_header (l_http_req, 'Content-Length', LENGTH (l_request));
       UTL_HTTP.write_text (l_http_req, l_request);
       DBMS_LOB.createtemporary (l_response, FALSE);
       l_http_resp := UTL_HTTP.get_response (l_http_req);
       BEGIN
          LOOP
             UTL_HTTP.read_text (l_http_resp, v_buffer, 32767);
             DBMS_OUTPUT.put_line (v_buffer);
             DBMS_LOB.writeappend (l_response, LENGTH (v_buffer), v_buffer);
          END LOOP;
       EXCEPTION
          WHEN UTL_HTTP.end_of_body
          THEN
             NULL;
       END;
       UTL_HTTP.end_response (l_http_resp);
       l_response := DBMS_XMLGEN.CONVERT (xmldata => l_response, flag => 1);
       -- Extract the soap envelope from clob. Issue because of the 32767 characters limitation
       SELECT    DBMS_LOB.SUBSTR (                  -- Problem here
                    l_response,
                      INSTR (l_response, 's:Envelope>', -1)
                    - INSTR (l_response, '<s:Envelope'),
                    INSTR (l_response, '<s:Envelope'))
              || 's:Envelope>'
         INTO l_response
         FROM DUAL;
          -- Parse the xml. Works fine once the above issue is fixed
          SELECT xt.nextstart
            INTO n_next_start_record
            FROM XMLTABLE (
                    xmlnamespaces ('http://www.w3.org/2003/05/soap-envelope' AS "s",
                                   'http://tempuri.org/' AS "data"),
                    's:Envelope/s:Body/data:QueryUpdateLogRecordsResponse/data:QueryUpdateLogRecordsResult/XObject.m_element/QueryResult'
                    PASSING xmltype (l_response)
                    COLUMNS nextstart NUMBER (20) PATH '@NextStart') xt;
       DBMS_OUTPUT.put_line ('NextStart ' || n_next_start_record);
    END;

  • Does anyone know how to use pages so you can export pdfs from the internet and automatically drag words from the document into the file name of the pdf (i.e., author, title of a scientific paper)

    Does anyone know how to use pages so you can export pdfs from the internet and automatically drag words from the document into the file name of the pdf (i.e., author, title of a scientific paper). For example, if I am downloading a paper by smith called "Surgery" that was published in 2002, it will automatically set the file name in the download to smith- surgery 2002. I have heard pages is smart enough to do this.
    thank you

    Pages can export only its own documents. They may be exported as PDF, MS Word, RTF or Text files.
    Pages can import (ie. Open a file as, or Insert a file into, a Pages document) documents in several formats, but won't rename the document as you describe. Documents that can be Opened (eg. Text, AppleWorks 6 WP, MS Word files) are converted to Pages documents, and retain their original names, with .pages replacing the original file extension. Files that can be Inserted (generally .jpg, .pdf and other image files) become part of the existing Pages file and lose their names.
    It may be possible, using AppleScript, to extract the text you want and to Save a Pages file using that text as the filename, but that would depend in part on being able to identify which text is wanted and which is not.
    How will the script determine where the author's name begins and where it ends?
    How will the script recognize the beginning and of the title, an decide how much of the title to use in the filename?
    How will the script recognize the year of publication?
    For papers published in a specific journal, with a strict format for placing each of these pieces on information, or containing the needed information as searchable meta data in the file, this might be possible. But it would require knowledge of the structure of these files, and would probably handle only papers published in a specific journal or set of journals.
    Outside my field of knowledge, but there are some talented scripters around here who might want to take a closer look.
    Best of luck.
    Regards,
    Barry

  • Bought an album off iTunes and two songs from the album will not download. Is there anyway to make them force load or anything along those lines?

    Bought an album off iTunes and two songs from the album will not download. Is there anyway to make them force load or anything along those lines? The Album is Sunshower-EP by UME and the songs are The Conductor and The Means.

    Hey ChandlerAdaway!
    I have an article here that can provide you some guidance for this issue. First, you will want to check for the songs in your past purchases list for your account, if you haven't already:
    Downloading past purchases from the iTunes Store, App Store, and iBooks Store
    http://support.apple.com/kb/HT2519
    If the songs cannot be found there, you will want to report an issue with your download by following the directions in this article:
    How to report an issue with your iTunes Store, App Store, Mac App Store, or iBooks Store purchase
    http://support.apple.com/kb/HT1933
    Thanks for coming to the Apple Support Communities!
    Regards,
    Braden

  • Is there a way to add/delete words from the predictive text ?

    Is there a way to add/delete words from the predictive text options that my iPhone is giving me? I'm using an iPhone 6 Plus with iOS 8.
    There are specific words that are popping up in the area right above the keyboard where it displays 3 word suggestions to choose from. This area seems to suggest words that I have typed before, I'd like to remove these words. Default seems to select the word in the center and when I'm typing fast on occasion it will recommend and select words that I don't need to use.
    Any help would be appreciated.

    999753 wrote:
    I know in 10g delete is added but in 9i i can only work with UPSERT( update, insert)
    is there a way to add delete into 9i MERGE? or somehow encapsulate it?
    Thank youI would suggest upgrading to a supported version of the database. 9i has been unsupported for many many years now.
    And to answer your question... to the best of my knowledge, there is no workaround. It works as it works.

  • When listening to music on ATV, it plays two songs from the playlist, then won't play any more.  You can see playlist and songs, but won't respond.  Any ideas?

    When listening to music on ATV, it will play two songs from the playlist, then no more.  You can see the playlist on songs, but it won't play.  Turned everything off, reset the network...nothing works.  Any ideas?  It used to play fine...for hours at a time.

    Hi.  Thanks for your reply.
    It is happening via "shuffle" in a playlist.  Always two different songs within the playlist.  You can see the next song on the screen that should, theoretically, be playing...but nothing happens.  Even when you push "play" on the remote, it doesn't receive the command. 
    I reset, then restored ATV.  Now, it doesn't limit itself to two songs, but does stop after a few.  The more trial and error I do, the more I believe the problem is with Homesharing in iTunes.  If I turn off Homeshare, then turn it back on, a couple more songs will play in the playlist.  I have shut down iTunes, restarted computer, completed updates, but nothing seems to change it.
    It used to play fine for hours on end, but has now found this glitch.

  • Using WQL "one-liner" to extract a value from the registry

    Can I extract a value from the registry using a "one-liner" WQL command?  Something like the following:
    Under root\default:StdRegProv
                   SELECT <Data> FROM "HKLM:SOFTWARE\Toto\Version"
    Please note that I am aware of how this is done via script.  The problem is that I'm using a management system (SCOM), which only allows me to supply a simple WQL query to perform my evaluation.
    Thanks,
    Larry

    Hi Larry,
    There have a specific forum to support the scripting related question, i sugges you ask in Scripting forum there will have more
    professional engineer will help you.
    The Official Scripting Guys Forum!
    https://social.technet.microsoft.com/Forums/en-US/9d5a7990-b975-488a-b7c0-6d866f29cf0a/change-mouse-scheme?forum=ITCG
    Best Regards,
    Please remember to mark the replies as answers if they help and unmark them if they provide no help. If you have feedback for TechNet Support, contact [email protected]

  • IPod updates 2 touch with the new software, and from the update they are not connected with Time phase. They appear calling but never they enter the calls. I make calls with other users and if one connects but in these two not from the update

    iPod updates 2 touch with the new software, and from the update they are not connected with Time phase. They appear calling but never they enter the calls. I make calls with other users and if one connects but in these two not from the update

    I use Firefox 95% of the time, and there's no problem with flash content (and I'm still at .55 - downloaded .64 last week but I haven't got round to making the change yet). At the rate they've been changing it recently it may well be out of date already .
    I've been trying it in Safari, too with no problems apart from You Tube, but that's because I'm blocking Google cookies.
    One thought does occur - if your Flash preferences are set to block all Local Storage, it may be that the problem site is trying to use Flash cookies (LSOs). I had this problem recently with the BBC iPlayer streaming content. Little Snitch notified an attempt to connect to a new URL - emp.bbci.co.uk.
    If I disallowed it, no streaming; when I allowed the connection, it still wouldn't stream, so I did a bit of digging and came to the conclusion that it was trying to set a Flash cookie. When I unblocked, lo and behold - streaming resumed as normal.
    I now have LSOs blocked in a more subtle way that lets the site think it's being set (but it ain't) and the streaming still works.
    As for permissions repair - always repair from local; the permissions on the original disc will have been superceded by updates and new installations of Apple software.
    DU needs to be reading the packages on the HD, not the (now out of date) install disc.
    The recurring repair messages are normal and don't mean anything's wrong. As long as the final message is 'repair complete' there's no need to dwell on it.
    http://support.apple.com/kb/TS1448 (for Leopard and Snow Leopard - Lion no doubt has it's own set).

  • Compare two results from the same table

    i have two results from the same table that i would like to compare. below is my query and the results i want to compare
    SELECT tblItemRoutingBOM.ItemRevID, tblItem.ItemID, tblItem.PartNum, tblItem.ItemName, tblItem.ManufacturerPartNum AS [Mfg Part#], tblItemRoutingBOM.Quantity
    FROM tblItemRouting
    INNER JOIN tblItemRoutingBOM ON tblItemRouting.ItemRoutingID = tblItemRoutingBOM.ItemRoutingID
    INNER JOIN tblItem ON tblItemRoutingBOM.ItemID = tblItem.ItemID
    WHERE tblItemRoutingBOM.ItemRevID in (61,70)
    as you can see i am returning two records using the where clause
    ItemRevID, ItemID, PartNum, ItemName, Manufacturer, Mfg Part#, Quantity
    61,121,331503,.233 Aluminum Sheet,,1
    70,121,331503,.233 Aluminum Sheet,,3
    now what i am looking for is to combine these two together into one row with the following added.  two columns for each QTY, QTY1 = 1 and QTY2 = 3 with a third column added that is the difference between the two QTY Diff = 2
    Any thoughts?

    Here are the two statements that i want to combine, results for each are attached
    SELECT tblItem.ItemID, Sum(tblItemRoutingBOM.Quantity) AS SumOfQuantity, tblItem.PartNum AS [Part #],
    tblItem.ItemName, tblManufacturer.ManufacturerName AS Manufacturer, tblItem.ManufacturerPartNum AS [Mfg Part#]
    FROM tblItemRouting
    INNER JOIN tblItemRoutingBOM ON tblItemRouting.ItemRoutingID = tblItemRoutingBOM.ItemRoutingID
    INNER JOIN tblItem ON tblItemRoutingBOM.ItemID = tblItem.ItemID
    INNER JOIN tblUnits ON tblItem.UnitID = tblUnits.UnitID
    LEFT JOIN tblManufacturer ON tblItem.ManufacturerID = tblManufacturer.ManufacturerID
    WHERE tblItemRoutingBOM.ItemRevID=61
    GROUP BY tblItem.ItemID,tblItem.PartNum,tblItem.ItemName,tblManufacturer.ManufacturerName,tblItem.ManufacturerPartNum
    SELECT tblItem.ItemID, Sum(tblItemRoutingBOM.Quantity) AS Quantity, tblItem.PartNum AS [Part #],
    tblItem.ItemName, tblManufacturer.ManufacturerName AS Manufacturer, tblItem.ManufacturerPartNum AS [Mfg Part#]
    FROM tblItemRouting
    INNER JOIN tblItemRoutingBOM ON tblItemRouting.ItemRoutingID = tblItemRoutingBOM.ItemRoutingID
    INNER JOIN tblItem ON tblItemRoutingBOM.ItemID = tblItem.ItemID
    INNER JOIN tblUnits ON tblItem.UnitID = tblUnits.UnitID
    LEFT JOIN tblManufacturer ON tblItem.ManufacturerID = tblManufacturer.ManufacturerID
    WHERE tblItemRoutingBOM.ItemRevID=70
    GROUP BY tblItem.ItemID,tblItem.PartNum,tblItem.ItemName,tblManufacturer.ManufacturerName,tblItem.ManufacturerPartNum
    114,11,55002,Pepsi Blue Cap,NULL,
    117,5,331501,Marigold Yellow For ABS,NULL,
    121,1,331503,.233 Aluminum Sheet,NULL,
    125,2,331504,Velvet Vinyl .008,NULL,
    114,33,55002,Pepsi Blue Cap,NULL,
    117,15,331501,Marigold Yellow For ABS,NULL,
    121,3,331503,.233 Aluminum Sheet,NULL,
    125,6,331504,Velvet Vinyl .008,NULL,
    my returned result should combine above with two extra columns (two extra columns because i have two results to combine)
    114, 11, 33, 22, 55002, Pepsi Blue Cap, NULL,
    117, 5, 15, 10, 331501, Marigold Yellow For ABS, NULL
    121,1, 3, 2, 331503, .233 Aluminum Sheet, NULL
    125, 2, 6, 4, 331504, Velvet Vinyl .008, NULL
    Columns go as such, ID, QTY1 (for 61), QTY2 (for 70), Diff (QTY1-QTY2), PartNum, ItemName, Mfg, Mfg Part#
    IF the results from one of those two are empty then i would see something like this
    114, 11, 0, 11, 55002, Pepsi Blue Cap, NULL,
    117, 5, 0, 5, 331501, Marigold Yellow For ABS, NULL
    121,1, 0, 1, 331503, .233 Aluminum Sheet, NULL
    125, 2, 0, 2, 331504, Velvet Vinyl .008, NULL

  • Hi, I just wanted to buy two iphones from the apple store usa with a lebanese credit card. Is it possible?

    Hi, I just wanted to buy two iphones from the apple store usa with a lebanese credit card. Is it possible?

    The US Apple Store accepts US-issued credit cards only. And they will ship only to a US address. Whether or not a physical Apple Store would accept a Lebanese credit card I don't know.
    Regards.

  • Compare two members from the same dimension in HFR

    Hi,
    Is it possibe to compare two members from the same dimension in HFR? The requirement is to compare Year and Week members from the same dimension. The Week date will be selected from POV. The corresponding Year date should be displayed in the report. Week dates are in the format W2008-03-07 and Year dates are in YTD2008-03-07.
    The dates are same except the preceding character.I am unable to compare these two. In my understanding there's no substring or replace functions in HFR.
    Kindly help. Thanks in advance.
    Regards,
    Uma

    Hi,
    How is your database structured? it may be possible to use the 'RelativeMember' function if it will always be the same number of steps between the 'W' member and the 'YTD' member, e.g. if your hierarchy is something like:
    Time
    .Weeks
    ..W2008-03-07
    ..W2008-03-10 etc.. for 52 weeks
    .YTD
    ..YTD2008-03-07
    ..YTD2008-03-10 etc.. for 52 weeks
    In your report select 'Current Point of View for Time' in one row/column and in the other use:
    RelativeMember set up as follows:
    Member: Current Point of View for Time
    Offset: 52
    Hierarchy: Time
    RelativeMemberList: Lev0, Time
    UseFirstDescendant: leave unselected
    Hope this helps
    StuartGame
    www.analitica.co.uk

  • I saw Corey Barker do a demo where he extracted an image from the background, then cleaned up the edge with a process that allowed him to de-fringe it.  I can't remember how he did it. And I can't find it.  Help!

    I saw Corey Barker do a demo a few weeks ago where he extracted an image from the background, then used a technique I was not familiar with to clean it up. You selected the area, then did something that involved going either to edit or select, then there was a dialog box that allowed you to dial it in, depending on the color background you were removing.  I can't find this anywhere. Can't remember.  It's driving me crazy!  If someone can help me find this, I would be very glad and grateful. Thanks!  Laura

    In technical support, sometimes you have to make educated guesses. I'm sorry that you were offended.
    iTunes does prompt when it is going to erase a device, and the message is clear.
    She said in her message that she was able to successfully sync the old ipad. This indicated to me that itunes wiping the data was not an issue, because either it had been setup at the apple store (in which case it doesn't actually wipe the ipad despite saying it will*) (*based on a single case I saw), or because the itunes media folder was migrated.
    Furthermore, my solution was to tell her how to backup her ipad (by either doing it manually, or as a last resort, by deleting the corrupt backup -- that she couldn't access anyway.)
    I got that last part of the instructions from the "Taking Control of your iphone" book which I found samples of when I did a google search for "corrupted backup itunes".
    She marked this as a solution, so it worked for her.

  • How to extract HTML page from the internet

    i am new to java, i wish to know how to extract Html page from the internet and also how to identify the differences between the images and text information?

    You can create a java.net.URL that points to the file you want to "extract" and read the HTML code (or what ever that file contains) from there using the inputstream given by URL.openStream().
    The difference between images and text... well, images are embedded in html using the img-tag. example: <IMG src="http://forum.java.sun.com/images/reply.gif" alt="Reply">. Attributes width, height, alt are sometimes left out and there may or may not be quotes around the values and everything is case insensitive... you'll be having hard time trying to parse the input so I'd suggest using existing parsers.
    What are you trying to do anyway? You can load a URL directly to a JTextEditorPane with the setPage(URL page) method...

  • HT3455 I got two receipts from the same tv show I bought. How's that even possible? I clicked the 'buy season ' button the first time and there was an error so I clicked it again, and now, I apparently paid for it twice!  Apple is such a rip-off!

    I got two receipts from the same tv show I bought. How's that even possible? I clicked the 'buy season ' button the first time and there was an error so I clicked it again, and now, I apparently paid for it twice!  Apple is such a rip-off!

    Wow.
    Over react much?
    Check your purchase history.  See if it actually charged you twice.
    If it did then contact itunes support for a refund.
    No need be so over dramatic.

  • Pages IPad: how to  'Find' occurrence of a word in a range starting from somewhere in the middle of document to the end. It seems that 'Find' feature always defaults to finding the word from the start of the document. Thanks

    Pages IPad: how to  'Find' occurrence of a word in a range starting from somewhere in the middle of document to the end. It seems that 'Find' feature always defaults to finding the word from the start of the document. Thanks

    Pages IPad: how to  'Find' occurrence of a word in a range starting from somewhere in the middle of document to the end. It seems that 'Find' feature always defaults to finding the word from the start of the document. Thanks

Maybe you are looking for