Search

Location:  Search Menu Bar

General

Using the industry standard, dtSearch, the Lexbe eDiscovery Platform (LEP) offers comprehensive, multi-index search methods to search for specific documents across an entire case using various search options.    

LEP offers search results generated not only from imaged/OCRed documents, which capture image-based text, but also generates results through native extraction, meaning text that is excluded from print and/or hidden/embedded text, and then converts the Native files into PDF and runs the documents through OCR.  Going a step further, LEP offers an additional index to search across translated documents, using a multi-lingual search index, as well as a foreign language concept search.  



LEP's search engine looks through the full text of all documents in a case and returns any documents containing the desired search terms. The search results page displays a list of documents returned from the search that can be viewed and coded from the same page or saved for later review.  The Search document window displays documents, ranked according to how often and how prominently the search terms appear in the text. The total number of documents found in the search is also displayed. 

The LEP search engine 
generates separate search index terms for the original and/or translated document.
 
The Search page has seven parts:
1) A text box for the search term(s);
2) Search options, containing choices for type of search to be run (natural language, Boolean, concept, fuzzy, stemming, phonic);
3) Filters to further narrow search results;
4) Fields to be displayed in the search window;
5) Delete documents;
6) Multi Doc Edit for mass tagging search results; and
7) Export options for moving search results into Excel.

Search Documents by Case or Across All Cases

LEP allows searching for Documents by Current Case or across All Cases In an Account.  Enter search terms in the search box (e.g. Invoice), separate multiple terms with spaces, and click Search or Enter. 


By default the main table will display the search for all documents in the Current Case only. To search across all cases in an account, choose the All Cases In Account option, click Search, and a new window will open and display results for all cases to which the user has access in LEP (see below).  
In order to view the details of this search in each case in the account, click the Select link for the case.  Click OK to confirm and access the selected case. Any document coding, custom fields, or other notes applied to search results in one case, will need to be created and mirrored in each individual case by the user. LEP will not do this automatically. Professional Services can assist in loading any custom fields or coding across an account if needed. 

How To Apply Searches (Search Box)

Enter specific keywords or search string commands in the search box and choose a custom search option from the drop down menu (i.e. All of the words, Any of the words, The exact phrase, or Boolean). Additionally, the user has an option to select a Concept Search, Fuzzy Search, Stemming Search or Phonic Search. More information on each type is contained below. 



Options

There are four types of search-term options available from the Options drop-down menu:


All of the words: Searches for all the specified words occurring in a document. A document will be returned in search results only if all the words are present.
Any of the words: Search results for any of the multiple keywords specified and present in the document.
The exact phrase: A document will be returned in search results only if the exact phrase is present in a document.
Boolean: Searches using a Boolean expression containing AND, OR and NOT. Boolean searches are discussed below.
Certain characters are invalid (i.e. "<" and ">") and will prompt an error.  

Advanced Boolean Searches

LEP supports advanced Boolean searches. A Boolean search consists of a group of words, phrases or macros linked by connectors such as AND and OR that indicate relationship or logic.

Boolean Search Examples
 Search Request                 Explanation
 apple AND pear both words must be present
 apple OR pear either word can be present
 apple w/5 pear apple must occur within 5 words of pear
 apple pre/5 pear apple must occur 5 or fewer words before pear.
 apple not w/5 pear apple must not occur within 5 words of pear
 apple and not pear only apple must be present.
 apple w/5 xfirstword apple must occur in the first five words
 apple w/5 xlastword apple must occur in the last five words

Boolean Search Tips
To use more than one connector, use parentheses to indicate the precise search criteria.  For example, apple and pear or orange juice could mean (apple and pear) or orange, or it could mean apple and (pear or orange).
 
In a Boolean search, there is no need to use special punctuation or commands to search for a phrase. Enter the phrase the way it ordinarily appears.  A phrase may be used anywhere in a search request. Example: apple w/5 fruit salad

Boolean Wildcards

A Boolean search word can contain the wildcard characters "*" and "?".

A ? in a word matches any single character and a * matches any number of characters. 

The wildcard characters can be in any position in a word. 
For example: 
appl* would match apple, application, etc. 
*cipl* would match principle, participle, etc. 
appl? would match apply and apple but not apples. 
ap*ed would match applied, approved, etc.

Please note that the use of the * wildcard character near the beginning of a word will slow searches somewhat.


Fuzzy Search

Fuzzy searching will find a word even if it is slightly misspelled. For example, a fuzzy search for "litigation" will find "liti3ation." Fuzzy searching can be useful when searching text that may contain typographical errors or for text that has been scanned using optical character recognition (OCR). Limit fuzziness selectively by using the % character. The number of % characters determines the number of differences dtSearch will ignore when searching for a word. The position of the % characters determines how many letters at the start of the word have to match exactly.  Examples:
ba%nana - Word must begin with ba and have at most one difference between it and banana.
b%%anana - Word must begin with b and have at most two differences between it and banana.

Deselect the fuzzy search checkbox if documents have not gone through OCR or if getting false hits on search returns.

Concept Search
 

Concept searching is a way to expand a search beyond specific keywords to include related concepts (synonyms and derivatives).  For example, a search in LEP without concept search enabled for the term "injury" would return documents that included the term injury and derivatives, like injuring and injuries. However, a concept search on "injury" would additionally return documents that mention trauma, abrasion, wound, etc. 


To enable Concept Search for a particular search, check the Concept Search box before clicking Search.

Stemming Searches

Stemming extends a search to cover grammatical variations of a word. A search for "fish" would also find "fishing." A search for "applied" would also find "applying", "applies", and "apply". This option enables stemming for all of the words in the search request. Stemming does not noticeably slow searches.

To add stemming selectively, add a ~ at the end. Example: apply~

The stemming rules included with dtSearch are designed to work with the English language. These rules are in the file stemming.dat. To implement stemming for a different language, or to modify the English stemming rules that dtSearch uses, edit the stemming.dat file. See the stemming.dat file for more information. 

Phonic Search 

Phonic searching looks for a word that sounds like the word being searched and begins with the same letter. For example, a phonic search for "Invoice" will also find "Investors", "Infection" and "Investment."

Noise Words

When a search request includes noise words, LEP will skip them and process the rest of the request as if the noise word matched every word in the search index.  For example, a search on the exact phrase "the Statue of Liberty" would return "the Statue of Liberty" in text in an indexed document, even though "the" and "of" are indexed and not searched. It would also return a hit on "of Statue the Liberty" or "xxx Statue yyy Liberty."

Below is a table with some noise words to take into consideration when forming search commands.
Begins with...Noise words
Aa, about, after, all, also, an, another, any, are, as, and, at
Bbe, because, been, before, being, between, but, both, by
Ccame, can, come, could
Ddid, do
Eeach, even
Ffor, from, further, furthermore
Gget, got
Hhas, had, he, have, her, here, him, himself, his, how, hi, however
Ii, if, in, into, is, it, its, indeed
Jjust
Llike
Mmade, many, me, might, more, moreover, most, much, must, my
Nnever, not, now
Oof, on, only, other, our, out, or, over
Ssaid, same, see, should, since, she, some, still, such
Ttake, than, that, the, their, them, then, there, these, therefore, they, this, those, through, to, too, thus
Uunder, up
Vvery
Wwas, way, we, well, were, what, when, where, which, while, who, will, with, would
Yyou, your

Sorting Search Results by the Built-in Doc and Custom Doc Fields

When searching for specific documents, optimize the search engine performance and sort the search results by Built-in Doc and/or Custom Doc Fields from the Filters option to the left of the search results window Sorting of any field or combination of fields is also available by exporting the desired fields as columns into Excel and then sorting using Excel sorting functions. 

Search for a Date Range (Boolean Option)

To search for a date range, type the range in the search box with "to" between the dates (i.e. 11-20-2008 to 11-30-2008). 



Build URL Arguments To Search Date or Date Range

The user may search a date range using the URL with optional arguments:
https://lexbe.ediscoveryplatform.com/Main/Search.aspx?q=date(Jan+10+2010+to+Jan+15+2010)
This URL argument may be constructed manually or from an Excel spreadsheet to automatically generate links for large numbers of search terms or to save terms to repeat searches.

Date Searches without Bounds
Indeterminate ranges are not supported.  To search for any date after or before a particular date, enter a bounded range with a maximal or minimal value for the bounds. The maximum value for a year is 2900, and the minimum value is 1000.  Example: date(jan 10 2006 to jan 1 2900) 
 
Combine a Date Range Filter with Search Terms or other Filters
A date range may be combined with a search term as follows with an AND join.
Type the following in the search box:
date(11-20-2008 to 11-30-2008) AND Smith
Or the following in the Browser URL:
https://lexbe.ediscoveryplatform.com/Main/Search.aspx?q=date(Jan+10+2010+to+Jan+15+2010)+AND+Black
 
The search results may be further limited with another filter (Select Filter from the left of the Search page) or by limiting the search to a filter.

No results found 

If there are no matching results, the search box will display the following:



Document Viewer -- Hits Tab

The Search is designed to open documents in Hits. Users can easily see where in the document the search terms are being found.  Click on the Text hyperlink next to each section of hits to jump to that place in the document.

Applying "WHERE" Logic To Fielded Data

LEP structures the user interface as a filter (similar to Excel) rather than as a SQL WHERE query as follows:

To code a General Field  (e.g., Doc Type = Email Body), from Browse or Search open the filter dialog; scroll to Doc Type and click into the box; the first ten entries will display for selection. Starting to type will limit or expand selection. This can be combined with a search.  To search, for example, for "Invoice" and filter on Doc Type = Email Body, it will return the documents similarly to an SQL query "Invoice WHERE Doc Type equals Email Body." The NOT operator is available on most (but not all) fields in the filter dialog (e.g. responsive, privilege, etc. and our custom dynamic coding fields). 

Search History

Search History records each term entered in the search engine specific to a user name.
 
Search Quick Links (Shared Function)

A list of all searches are automatically saved in the database whenever the user enters a keyword and clicks on Search.  Search records are preserved on the list view under the Recent and creates filter hyperlinks to open a specific set of documents. The default link titles are under the Recent Search YYYY-DD-MM format.  To rename, click the Edit hyperlink. The user may also pin and share the searches, delete existing searches or Cancel. 



Search Quick Links section is divided into three subsections:

Shared. To share or unshare searches with other users in the current case, click on the Edit link.  When the Shared/Pinned Searches dialog box appears, check or uncheck the shared icon checkbox. This option will share searches with all current users that have access to the case.
Pinned. Click on the Edit hyperlink, select a recent search by title and use the checkbox under the pin symbol to pin searches. Click OK to save the changes or Cancel. The pin icon will move the selected search to the Pinned section. 
Recent. Shows the search history applied.  
The screen displays the five most recent searches.  Click the Edit hyperlink for a full list of searches.

To increase search query speed, LEP does not index the Built-In or Custom Doc fields data.  Those fields can be filtered to locate a more specific set of documents.

Warning Message On Very Large Cases

The search engine stops a search after a predefined period and returns the results found up until that point. The above warning will display with the search results if this occurs.  The search engine also stops a search after returning 250,000 results.

Try the following to enable the search to complete across the entire database:
Make the search terms more specific.
Uncheck (if checked) the search options for Concept Search, Fuzzy Search, Stemming Search, and Phonic Search, as these slow down search completion.
Switch from All of the Words or Any of the Words to Exact Phrase or Boolean.
Apply a Filter before starting a search.

ESI Culling

The process of culling (reducing or filtering) electronically stored information (ESI) is designed to defensibly reduce a large collection of documents to a smaller set of potentially responsive documents requiring further review.  Depending on how documents were collected and other factors, culling can reduce data sets by 90% or more (if broadly collected) and thereby significantly reduce eDiscovery costs.  Culling can be done by Custodian, by date range, and by text.  See ESI Culling for more information.  

Lexbe Professional Services is available to assist with any culling needs or keyword searching.  Contact your sales consultant for a quote, or email professionalservices@lexbe.com to speak to a Professional Services representative about any questions related to searching or culling.