If you just ran a search and didn’t get the results you were hoping for, try the following techniques.

Change to the English page:That’s any of the following:

@ + # - _ / \ . , : = > * ? ( ) [ ] $

Our search engine sometimes interprets these characters as “logical operators” rather than literal strings. For example, you might intend a : character in your query to be a colon. However, search may be interpreting it and the text around it as a field query. That will change what your query means to the search engine, and may affect the results.

In order to avoid this, do one of the following:

1. Enclose your query in quotes, such that binarytuplewriter::writetuple becomes “binarytuplewriter::writetuple”.
2. Enclose your query in the “no syntax block” notation, such that binarytuplewriter::writetuple becomes <@- binarytuplewriter::writetuple -@>. (Please note the spaces separating the notation from the query.)

This technique is often effective when you’ve pasted an error message or log file entry straight into search.

That could be a pathname that’s unique to your local computer, a server name that’s unique to your environment, or a user name that’s unique to your organization. Error messages sometimes contain such personalized information, such as:

Errors occurred while trying to load the workbook “c:\users\mikeb\mike’s files\reports\sales\april2017\US_sales.twbx” …

A query like this will never return a relevant article or community post. That’s because our writers cannot know what your pathname, server name or user name will be.

The solution is to pick the longest string from the error message likely to return a relevant document. In the case above, try searching on “Errors occurred while trying to load the workbook”. This is more specific than just “Errors occurred”, and is likely to save you time when scanning the search returns for the material you need. At the same time, it excludes the personalized information that will cause the search to fail.

Use the advanced filters options that are available on the search results page. Do not hesitate to explore the different types of content available on the left side of the screen.

Our search engine has been optimized to enable effective searching in Chinese, Japanese and Korean. In order to do that, it uses word tokenizers to identify and separate expressions into individual groups of inseparable characters (words). Search indexing uses these separate words (rather than the individual characters) to assign relevance ranking, and to perform retrieval.

A more detailed description of how this works, and how to use it, is available at Chinese Japanese Korean Language Improved Relevance , a document provided by our search provider, Coveo. Please note that PDF downloads are available for Simplified Chinese, Japanese, Korean and Thai translations of this document.