Searching (High)

*Using keyword search- Overview* The keyword search index details page contains the following fields:

-Name (display name of the keyword search index) -Order (position of the index in the list. lowest-numbered at top, highest-numbered at bottom. ones with same value are sorted in alphanumeric order. order can be any integer (positive or negative) NO DECIMALS ALLOWED. -Active (determines whether the index should be activated or deactivated. Yes (activated) No (deactivated)

*Using dtSearch search- Using dtSearch syntax optons* Date recognition searches for strings that appear to be dates. It uses English-language months, including common abbreviations and numerical formats. How do you search for a date with this date recognition?

-enter a date expression between parens in string date() e.g. date(jan 10 2006) -can also do a range e.g date(jan 10 2006 to jan 20 2006)

*Using dtSearch search- Creating a dtSearch index* How many characters of a string does dtSearch index?

32. Within a field, dtSearch truncates any string longer than 32 characters that doesn't contain a space character. It indexes only the first 32 characters of the string.

*Using dtSearch search- Creating a dtSearch index* An incremental build updates an index after adding or removing documents. During an incremental build, is the index still available for searching?

During an incremental build the existing index remains available for searching, but changes to the index are not reflected in search results until the incremental build is complete.

*Using dtSearch search- Using dtSearch syntax optons* Credit card number recognition searches for any sequence of numbers that matches the syntax for a valid credit card number. How do you use this syntax?

Enter the credit card number between the parens in creditcard() e.g. creditcard(1234*(

*Using dtSearch search- Creating a dtSearch index* (T/F) The incremental build process detects when documents have been added or removed for an index' data source search as well as changes have been made to the fields to be indexed, such as when the extracted text of an indexed document has been changed or populated for the first time.

False -It can only detect when documents have been added or removed from the data source -It CANNOT detect when changes have been made to the fields to be indexes

*Using dtSearch search- Creating a dtSearch index* Network problems can slow down your dtSearch builds. If a dtSearch manager or worker agent encounters a network-related error during the build process, how many retry attempts will it execute?

It will execute up to three retry attempts at 30 second intervals.

*Using dtSearch search- Creating a dtSearch index* When creating a dtSearch index, what is the best practice for which fields to index?

It's best practice to only index the fields you want to search. Move all other fields to the Fields (Required) left column. Typically, you only index the extracted text field if you're searching the body of emails.

*Using dtSearch search- Using dtSearch syntax optons* What does auto-recognition do?

Provides the ability to search for various date formats, e-mail addresses, and credit card numbers. It can dramatically affect indexing and search performance when activated.

*Using keyword search- Overview* Define Keyword search

Keyword search (or SQL index search) is Relativity's default search engine. You can use a keyword search to query a full text index.

*Using dtSearch search- Creating a dtSearch index* If you click "cancel build" this aborts the indexing thread, leaving the index in an unstable state. What will Relativity do with these indexes?

Relativity deletes these indexes from the population table and gives them an inactive status. You can't search against an index with an inactive status until you run a full build. Canceling also deletes the index files in the file share.

*Using keyword search- Overview* If you apply item-level security to a search index, can users run any public saved searches built on that index?

No, users can't run any public searches on that index with item-level security applied and will get an error. We recommend leaving the index unsecured and instead applying security to a Search indexes tab or individual saved searches.

*Using keyword search- Overview* When you click Sort By Rank when using keyword search in the search panel, what does this do?

Orders results in order by relevance. The most relevant documents are listed at the top of the result set.

*Using dtSearch search- Using dtSearch syntax options* What do we strongly recommend using to group expressions and control which ones are evaluated first??

Parentheses. For example, the parenthetical grouping in the following search string directs dtSearch to evaluate the AND expression before the OR expression: grape or (apple AND pear)

*Using dtSearch search- Creating a dtSearch index* Since the incremental build process can't detect when changes have been made to the fields to be indexes (e.g. adding or changing extracted text), what must you do in order to capture changes to an indexed field.

You must perform a full build in order to capture those changes to any indexed fields

*Using dtSearch search- Using dtSearch syntax optons* When must you use double quotes in dtSearch?

You must use double quotes when searching for exact phrases than contain words that are reserved as dtSearch operators, such as boolean connectors AND e.g. clear and present danger (returns docs with the word clear and the phrase present danger) "clear and present danger" (returns the exact phrase clear and present danger) but also remove and from the noise word list

*Using dtSearch search- Using dtSearch syntax optons* What dtSearch behavior ensures consistent result sets when querying with the AND NOT operator?

The dtSearch engine combines into a single pool the text for all fields identified for inclusion in an index. A search string using the AND NOT operator queries the index that includes the combined text from all indexed fields, rather than querying the content of individual fields.

*Using dtSearch search- Creating a dtSearch index* The sub-index fragmentation threshold default value is 9 and must be equal to or greater to one. What value determines the default value?

The dtSearchDefaultSubIndexFragmentationThreshold instance setting value

*Using keyword search- Overview* Which long text and fixed-length fields are included in the keyword search index?

The long text and fixed-length text fields included in this index vary by workspace.

*Using dtSearch search- Using dtSearch syntax optons* How does the AndAny operator work?

The words before the AndAny connector are required search terms, and the words after the AndAny connector are optional.

*Using dtSearch search- Using dtSearch syntax optons* dtSearch indexes are case insensitive by default. All characters in a dtSearch index are normalized to lowercase. What would you do if you need to search for an exact acronym like ACT?

Then you must build a case-sensitive dtSearch index.

*Using keyword search- Overview* As an alternative to the Ryan NOT Will keyword search conundrum above, what is an alternative?

You can use the AND NOT operator in a dtSearch as an alternative approach to this type of keyword search.

*Using keyword search- Overview* (T/F) when you search on the term Pop-up, both of these below will return: Pop.up Pop--up

True (hyphens and dashes) When a search phrase includes a hyphen or dash, the query returns results that include terms containing other punctuation marks.

*Using keyword search- Overview* (T/F) Single uppercase and lowercase letters are default stop words, so you cannot query on them with a keyword search. Each single letter [A-Z and a-z] is considered a stop word.

True (single letters as stop words section) However, you can query on a capital letter followed by a period. The SQL search engine assumes this is an abbreviation. For keyword searches, this functionality is available only on queries in the English language. It does not apply to lower case letters followed by a period, which are still considered stop words.

*Using keyword search- Overview* (T/F) While the keyword search offers fewer options than other Relativity searches, it uses an index that's automatically populated, reducing maintenance and ensuring all required document fields are indexes.

True as hell.

*Using keyword search- Overview* (T/F) Search terms with accented letters are recognized and return keyword search results.

True as hell. gray note in doc.

*Using keyword search- Overview* (T/F) You can use Boolean operators (AND, OR, NOT) in keyword searches, as well as quotation marks for exact matches, asterisks (*) for wildcards, and other features.

True!

*Using dtSearch search- Creating a dtSearch index* (T/F) The Compress Index button only runs compression against sub-indexes that have a fragmentation level greater than zero.

True! And cancelling compression returns the index to its original fragmented state before compression began.

*Using keyword search- Overview* (T/F) Single digits 0-9 are default stop words, so you cannot query on them with a keyword search. Relativity doesn't return the expected results if you attempt to query on a single digit.

True! (single digits as stop words section) Use dtSearch to query on a specific number or letter. However, you can use a keyword search to query on whole numbers greater than 9. You can search on more than one digit, such as 09. While these digits may be used to represent a specific numeric value (such as 9), they are not considered single digits, and can be used in a keyword search.

*Using dtSearch search- Creating a dtSearch index* (T/F) Cancelling an incremental build returns the index to its previous state.

True.

*Using dtSearch search- Creating a dtSearch index* (T/F) Once compression is complete, the system automatically replaces the old sub-indexes with the defragmented sub-indexes. Duplicate sub-indexes are removed when the Case manager agent runs.

True.

*Using dtSearch search- Using dtSearch syntax options* (T/F) Fuzzy searching uses term length and fuzziness level to decide how many % characters to add. This is not a straight level to character match. This means a level 7 fuzziness search doesn't necessarily mean up to 7 additional characters return.

True.

*Using dtSearch search- Using dtSearch syntax options* (T/F) If you make any symbol a searchable character in your dtSearch index and then build an index on a long, uninterrupted search string, such as file path, dtSearch truncates the string after the 32nd character.

True.

*Using dtSearch search- Using dtSearch syntax optons* (T/F) When using Boolean operators in a proximity search, noise words are included. Although the noise words are not searchable, they are still counted in the proximity search.

True.

*Using keyword search- Overview* Stop words at the beginning or tail end of a keyword search string are ignored.

True.

*Using dtSearch search- Using dtSearch syntax optons* (T/F) Keyword search won't return the same results as dtSearch when the NOT operator is used to query across multiple fields.

True. A keyword search is a SQL full text search, which queries individual fields. Whereas the dtSearch engine combines into a single pool the text for all fields identified for inclusion in an index.

*Using dtSearch search- Using dtSearch syntax options* (T/F) The Fuzziness Level menu (that you can adjust in the Dictionary search or Saved Search) is independent of the fuzziness (%) character that you can enter in the textbox.

True. A search for appl% without a Fuzziness Level setting may return documents containing apple or apply, since these terms have the stem appl and differ by one character.

*Using dtSearch search- Using dtSearch syntax optons* (T/F) The colon (:) is not used by Relativity but is still considered a syntax term by the dtSearch index.

True. As such, the symbol cannot be made searchable by modifying the alphabet file.

*Using dtSearch search- Using dtSearch syntax optons* (T/F) A credit card number is recognized regardless of the pattern of spaces or punctuation embedded in the number.

True. Can be 1234-5678-1234-5678 1234567812345678 1234 5678 1234 5678

*Using keyword search- Overview* (T/F) the at sign (@) is ignored in a keyword search, when it is used at the beginning of a query.

True. For example, if you search a domain name, the same number of documents return whether you include or exclude @.

*Using keyword search- Overview* (T/F) If a keyword search string containing stop words is surrounded by double quotes, then the stop words' positions in the string are taken into account when the query is executed.

True. However, only the positions of any intervening stop words are taken into account, not the stop words themselves.

*Using keyword search- Overview* (T/F) Stop words used in keyword searches are ignored if the search string is not surrounded by double quotes.

True. In a search for the phrase. sun on my head, both on and my are ignored. The result is that the words sun AND head are queried without respect to proximity. Thus, any documents that contain both the words sun and head will be returned.

*Using keyword search- Overview* (T/F) Keyword searches do not support the use of wildcards at the beginning of a word.

True. Keyword searches are SQL index searches run on the Microsoft SQL Server, which does not support leading wildcards in full text searches.

*Using dtSearch search- Using dtSearch syntax optons* (T/F) The underscore (_) is not recognized as a space by default.

True. Verify that a given character is defined as causing a word break before using it as a space in dtSearch

*Using dtSearch search- Using dtSearch syntax optons* (T/F) In dtSearch, you can use the NOT operator by itself at the beginning of a search expression.

True. e.g NOT pear returns all docs that don't contain pear e.g NOT (apple w/5 pear) returns all docs that do not contain the world apple within 5 words of pear

*Using dtSearch search- Using dtSearch syntax options* (T/F) The number of (%) used indicates how many characters in the search term dtSearch ignores when it runs the query. The position of the % indicates the number of characters from the beginning of the term that must match exactly.

True. e.g app%ly matching words must being with app and differ from apply by only one character e.g a%%pply matching word must begin with a and differ from apply by only two characters

*Using dtSearch search- Using dtSearch syntax optons* (T/F) When the NOT operator appears in the middle of a search expression, it must be used in conjunction with either AND or OR.

True. e.g. apple OR NOT pear returns all docs that contain the word apple or those that do not contain the word pear

*Using dtSearch search- Using dtSearch syntax optons* (T/F) There is no rule that requires double quotes around a phrase of any number of words.

True. e.g. search string is apple grape banana, returns exact phrase apple grape banana

*Using dtSearch search- Creating a dtSearch index* (T/F) You can search against the original (uncompressed) dtSearch index while compression is in progress.

True. for the Compress Index button- this compresses the dtSearch index returning all sub-indexes with a fragmentation level greater than zero to a fragmentation level of zero.

*Using keyword search- Overview* (T/F) These 2 search strings below will return the same records "sun on my head" "sun my on head"

True. this is because keyword search evaluates both search strings as a query for the phrase sun ABC XYZ head, where ABC and XYZ represent any two words, not just stop words.

*Using dtSearch search- Creating a dtSearch index* (T/F) The searchable set for a dtSearch source may use a dtSearch or Relativity Analytics index.

True. But make sure the index is active.

*Using dtSearch search- Using dtSearch syntax options* Stemming returns documents containing grammatical variations of a root word. (T/F) Stemming is limited to English only.

True. English only

*Using dtSearch search- Creating a dtSearch index* (T/F) The incremental build process copies each sub-index that requires modification, updates the copy, then replaces existing sub-indexes with the updated copies. Duplicate sub-indexes are removed when the Case manager agent runs.

True. I just copied this, didn't know how else to get the info. in here.

*Using dtSearch search- Using dtSearch syntax options* (T/F) Do not remove these Unicode characters from your alphabet file (referring to the space characters)

True. gray note says not to

*Using dtSearch search- Using dtSearch syntax optons* (T/F) Do not confuse the parentheses function for order of preference with the double quotes function.

True. this is a note.

*Using dtSearch search- Creating a dtSearch index* (T/F) accented characters like á or ñ are converted to the unaccented versions, a or n (T/F) If you searching for the term fröhlich, searching that term as fröhlich or frohlich would both return the hit.

True. to first True. to second. However, highlighting in the Viewer may not display both variations. Because Relativty (by default, and in this version I believe- pretty sure there is a recipe out there now) uses only accent-insensitive indexes, some characters are translated to the base character, which causes those characters and any terms containing those characters to be deduplicated in a Search Terms Report

*Using dtSearch search- Using dtSearch syntax options* If your dtSearches aren't returning the expected results, what should you ask your admin about adjusting?

about maybe adjusting the noise word list or the alphabet file

*Using keyword search- Overview* If your search string is Ryan NOT Will will it return this doc below? (if you want to query for email messages that have Ryan as the author, but do not have Will as the recipient) document: AS00001 OCR: From Ryan To: Will Recipient: Will Author: Ryan

yes because the Author field has Ryan but not Will, so this doc is returned. it's searching field by field. this would unexpectedly return the document

*Using dtSearch search- Using dtSearch syntax options* There are default noise words yo. Soo if you're searching for a phrase that contains a term in the noise word list..what will you need to do??

you need to remove that term from the list and rebuild your index

*Using dtSearch search- Creating a dtSearch index* The Swap Index function updates anything in the Views table, which affects what?

batches, saved searches, nested searches, etc.

*Using keyword search- Overview* What will both of these queries without quotes actually search for? sun on my head sun on head

both will search: sun AND head

*Using dtSearch search- Using dtSearch syntax optons* Doc 1: Apple Doc 2: Apple, Grape, Pear Doc 3: Grape, Pear When you search apple, which documents return?

doc 1 and doc 2

*Using dtSearch search- Using dtSearch syntax options* dtSearch includes built-in search words xfirstword and xlastword. you can use these terms to limit a search to the beginning or end of a file. e.g does apple w/10 xlastword, search for apple within 10 words or 11 words of the end of the document?

doc says 11 but I'm really confused? wouldn't it just be 10? For example, apple W/10 xlastword searches for apple within 11 words of the end of a document.

*Using dtSearch search- Using dtSearch syntax optons* Doc 1: Apple Doc 2: Apple, Grape, Pear Doc 3: Grape, Pear When you search apple AndAny pear, which documents return?

docs 1 and 2

*Using dtSearch search- Using dtSearch syntax optons* dtSearch recognizes numeric strings as dates, as long as it can be interpreted as a valid date. includes US and UK formats. What does dtSearch do in the case of ambiguous dates, such as 01/05/10?

dtSearch defaults to MM/DD/YY if the date contains words dtSearch converts the words to a numeric value to help interpret the date. e.g 30 must be a day and not a month, and 2015 must be a year (not a day or month)

*Using dtSearch search- Using dtSearch syntax optons* What is the dtSearch evaluation for the following search string? apple AND pear OR grape

pear or grape is evaluated first then and apple is evaluted second

*Using keyword search- Overview* Certain punctuation marks are treated as stop words by default, so you cannot query on them in a keyword search. They include:

period (.) dash (-) colon (:) semicolon (;) slash (\,/)

*Using keyword search- Overview* Where can you run a keyword search from?

search bar search panel (can sort by rank here and add conditions) search browser (can sort by rank here and add conditions)

*Using dtSearch search- Using dtSearch syntax optons* You can use the AND NOT operator to develop queries for documents that include the first expression but not the second expression. If you want to use this to get results that DON'T contain this document what would the search string be? OCR- From: Ryan To: Will Recipient- Will Author- Ryan

search string: Ryan AND NOT Will

*Using keyword search- Overview* What will these queries with quotes actually search for? "sun on my head" "sun on head" "sun on my head and" "and sun on head"

sun [any word] [any word] head sun [any word] head sun [any word] [any word] head sun [any word] head

*Using dtSearch search- Creating a dtSearch index* What does Swap Index do?

swaps your index with a replacement index in order to use its resources while your index builds or is inactive or disabled for any reason. This enables you to keep searching while your primary index experiences downtime. You can only select indexes in the Replacement Index with an Active Status. The index swap to doesn't automatically run an incremental update.

*Using dtSearch search- Using dtSearch syntax optons* What is the precedence, or order of evaluation, for OR and AND expressions with dtSearch?

By default, dtSearch evaluates OR expressions before AND expressions: A AND (B OR C) Remember!! Unlike dtSearch, the order of precedence for a keyword search evaluates AND expressions before OR expressions (A AND B) OR C

*Using dtSearch search- Creating a dtSearch index* What type of fields might have text associated with them that is not visible in your document views?

File type fields, linked fields, and HTML enabled fields. This includes the system FileIcon field, which is populated with the original file name upon import.

*Using dtSearch search- Using dtSearch syntax options* What are noise words?

The noise words are excluded in a dtSearch index to improve query performance and prevent unnecessary index growth. Commonly used words such as AND, THE, WILL are ignored when you run a query.

*Using keyword search- Overview* When a search string does not include parentheses, what is the order of precedence for a keyword search?

The order of precedence for a keyword search evaluates AND then OR expressions. For example, the search string A AND B OR C is evaluated as (A AND B) OR C.

*Using dtSearch search- Using dtSearch syntax optons* (T/F) You can also use NOT in a proximity search as illustrated by the NOT W/N (NOT Within N words) operator.

True

*Using dtSearch search- Using dtSearch syntax options* What is fuzzy searching and what is the character used for it?

Used to return documents containing spelling variations of a specified term. You may want to sue this when querying documents that contain misspelled terms, typographical errors, or have been scanned with OCR. The percent sign (%) is the character used for fuzzy searches.

*Using dtSearch search- Using dtSearch syntax optons* Email address recognition searches for text with the syntax of a valid email address. You can search for a specific email address regardless of the alphabet settings for "@", "." or other punctuation. Can also use word listing functions to enumerate all addresses. Must inclue the * or ? wildcard to enumerate all emails in a doc collection. How do you search using this?

e.g mail([email protected]) returns exact email address [email protected] mail(sa*@example.com) - returns variations of the email address [email protected]; [email protected]

*Using dtSearch search- Using dtSearch syntax optons* Which Boolean operators does the dtSearch engine support to connect multiple phrases or terms in a single search expression?

includes AND, OR, and NOT

*Using dtSearch search- Creating a dtSearch index* What do you need to do before you create a dtSearch index?

Before you begin, you need to create a saved search that includes the fields that you want to include in the index.

*Using dtSearch search- Creating a dtSearch index* What is the sub-index fragmentation threshold?

Determines the fragmentation level at which the system automatically compresses a dtSearch sub-index during an incremental build. An incremental build automatically compresses any sub-index equal to or greater than the fragmentation threshold.

*Using dtSearch search- Creating a dtSearch index* What does the Show Detailed Status pop-up show?

Doc count- total number of docs in the index Index Size (in bytes) Created Date Updated Date Last Build Duration (how long the last build took in hours, minutes, and seconds) Refresh page- show the index's current build status.

*Using dtSearch search- Using dtSearch syntax options* How do you make a special character searchable?

Edit the alphabet file. delete the character from its current category (hyphen, spaces, etc.) enter the character you want to make searchable 4 times, separated by spaces under Letters. original letter, upper case, unacccented. YOU MUST ALSO BEGIN WITH A SPACE! THEN YOU NEED TO PERFORM A FULL BUILD ON THE INDEX.

*Using dtSearch search- Creating a dtSearch index* If you wanted to set a new default value (normally 250,000, minimum is 1000) for the sub-index size, what can you do?

Edit the the dtSearchDefaultSubIndexSize instance setting

*Using keyword search- Overview* (T/F) You can start key word searches with the NOT operator, or use it with the OR operator.

False as hell. Do NOT start key word searches with the NOT operator, or use it with the OR operator. For example, these searches are invalid: -not wired -wired or not magazine

*Using keyword search- Overview* (T/F) If you perform a keyword search with multiple terms, documents where those terms exist in separate fields will still return.

False!! If you perform a keyword search with multiple terms, documents where those terms exist in separate fields WON'T return.

*Using keyword search- Overview* (T/F) single uppercase and lowercase letters followed by a period are not considered stop words.

False. An uppercase letter followed by a period is not a stop word. For queries in English, SQL assumes it's an abbreviation A lowercase letter followed by a period is still considered a stop word

*Using dtSearch search- Using dtSearch syntax optons* (T/F) Credit card recognition would never find additional invalid credit card numbers.

False. Credit card issues use numerical tests to exclude sequences of numbers that aren't valid credit card numbers. Since these tests don't detect all invalid numbers, the feature for credit card number recognition may find additional invalid numbers.

*Using dtSearch search- Creating a dtSearch index* (T/F) Deactivate Index removes the index from the database.

False. it deactivates the index and removes it from the "search with" menu in the documents tab (but not from the database)

*Using dtSearch search- Using dtSearch syntax optons* (T/F) Unterminated date ranges are supported with date recognition.

False. To search for any date after or before a particular date, enter a bounded range with a maximal or minimal value for the bounds. max value is 2900, minimum 1000 e.g date(jan 10 2006 to jan 1 2900)

*Using dtSearch search- Using dtSearch syntax options* The alphabet file determines which characters are treated as text, which cause spaces, which cause word breaks, and which are ignored. What are the categories in the alphabet file?

LETTERS: all searchable characters which are all alphabet characters (a-z and A-Z) and all digits (0-9) HYPHENS: characters removed during index creation. for example "First-Level" becomes two separate words in a dtSearch index SPACES: characters that cause a word break. e.g the period is indexed as a space char. by default. so U.S.A is 3 separate words: U, S, and A. Values listed as \## are unicode characters IGNORE:characters that are disregarded in processing text. For example, if you classify the period as ignore instead of space, then dtSearch would process U.S.A as one word, USA.

*Using dtSearch search- Creating a dtSearch index* Selecting the index from the drop-down and clicking OK completes the index swap. Can you reverse the swap results in the current dialog box?

No you can't reverse the swap results. You must close this swap and run again to swap back or swap another time. Swapping is useful in very limited cases, if you are doing a full rebuild on a very large index for example.

*Using dtSearch search- Using dtSearch syntax options* What is phonic searching (#) ?

Phonic searching returns documents containing words that sound like the word you're searching for and begins with the same letter. The # is added to the front of the word. #pear- would also find pair and pare you can use this shit in Dictionary searches too

*Using dtSearch search- Using dtSearch syntax options* What does the alphabet file determine?

The alphabet file determines how characters and spaces are handled in a query.

*Using dtSearch search- Creating a dtSearch index* (T/F) If you search for long, uninterrupted strings that have no spaces or word breaks, such as when you've made a character searchable, dtSearch truncates the string after 32 characters and inserts a wildcard.

True

*Using dtSearch search- Using dtSearch syntax options* (T/F) The underscore (_) isn't recognized as a space by default. Verify that a given character is defined in the [Spaces] section before treating it as a word break in a dtSearch.

True

*Using keyword search- Overview* What's the best way to draft queries outside of Relativity, to prevent adding characters or formatting that might return unexpected search results?

Use a plain text editor such as Microsoft Notepad to prevent adding characters or formatting that might return unexpected search results.

*Using dtSearch search- Creating a dtSearch index* When does the "Activate this index upon completion" box display?

When you click "Build index: Full" you get a confirmation pop-up box. the box is checked by default. With it being checked, you don't have to actually click the activate button. Full build confirmationClose Are you sure you want to perform a full build? Any searches running against this index will be canceled Activate this index upon completion

*Using dtSearch search- Using dtSearch syntax optons* Doc 1: Apple Doc 2: Apple, Grape, Pear Doc 3: Grape, Pear When you search apple AND pear, which documents return?

only doc 2

*Using dtSearch search- Creating a dtSearch index* When does the system automatically compress a sub-index during an incremental build?

only when the sub-index fragmentation level is equal to or above the sub-index fragmentation threshold value. The incremental build process doesn't update the index for documents currently in the index with modified text.