Building a search query
Last updated
Last updated
Building a functional search query requires functions and keywords, as well as to add synonyms and select search filtres options. This is all possible from your query page, as shown here-under.
Let’s focus on some specific parts of the query
Document Search: This feature allows users to combine functions and keywords to find relevant literature, including papers and patents. Detailed instructions on how to perform an effective search are provided below.
Technology Search: This feature enables users to use natural language queries to find the most relevant technologies and literature (only scientific papers). Detail information can be found in the page: Technology Search.
IgorAI provides word auto-completion when searching. For instance, if you write Recycl, it will automatically also search for Recycl-ing, Recycl-ed, etc.
For precise searches, use quotation marks around keywords. For example, when searching for Aluminum as "Al," it ensures only exact matches without variations like Alabama or Almost are found.
When adding a function, combine a verb and a noun as the action-object pair. IGOR^AI will disregard other combinations but will look for grammatical variations. For instance, inputting "Inspect Weld" will include variants like "Has been inspected" to enhance search accuracy.
Researchers often use different terminology. Therefore, In search, using synonyms is crucial for capturing all results. It ensures broader coverage without compromising result quality. Synonyms can impact results in two ways:
To increase results, add more synonyms if needed.
To reduce noise, identify problematic synonyms and disable them as necessary.
You can manage synonyms in IGOR^AI through various options in the synonym management window, where you can:
Manually add synonyms
Include AI-generated synonyms
Access previously used historical synonyms
Disable or delete synonyms as needed
There are three options to include/exclude search terms in your query.: Must Match, Should Match and Should Not Match. You can use them to include, score or exclude search terms.
Sometimes, you might be using internal jargon, or words that are not used in sciencitif writing. Similarly, you can have words creating too much noise. Verifying the amount of search terms will help you to understand what those words are, but will also help you to define the right setting of Must/Should/Must not match.
Example:
For this query, "Inspect Weld" and "Metal" are set as Must Match keywords, resulting in 127 papers showing both the function and keyword together. However, the term "Offshore" appears only 3 times within these 127 results. These numbres are visible on the right of the keyword.
To enhance the search, additional synonyms for "Inspect Weld" and "Offshore" should be included. By incorporating 3 synonyms for "Inspect" and 2 for "Offshore," the Must Match results increase to 627 papers, with 70 focusing on "Of fshore." This refined query provides a more comprehensive starting point for further investigation.
To connect a query to a specific study or entity for ongoing use, simply type the study or entity name in the top bar. This allows you to link the query to the associated research or project for easy reference and continuity.
The filters for searching in science and patents are slightly different.
In science searches, filter by publication date, while patents can be filtered by publication or filing date.
Both allow narrowing down results by affiliations or assignees.
In science searches, choose from various document types.
In patents, include or exclude patents based on the country of origin by using country codes like US, EP, CN, etc.
