How Mergeflow's Boolean query syntax works
does not use brackets, and searches are
case insensitive (iot = IoT = IOT = iOT).
Using quotation marks will search for an exact match. For instance, “sensor network” matches documents containing “sensor network”. It does not match documents containing “sensor networks”, “sensor”, or “network”.
When you use queries consisting of more than one word in the start search page, Mergeflow assumes you mean exact matches. So...
...becomes "machine learning":
If you don't want to use exact matches, you can click on the green filter and remove the quotation marks (don't forget to hit 'return' when you're done):
You should also use quotation marks if you want to search for terms that contain brackets. This could be names of materials, such as benzo(k)fluoranthene, for example:
When you search for
machine learning AND power harvesting...
...you are searching for
documents that contain both "machine learning" and "power harvesting" (not for documents containing "machine learning" plus documents containing "power harvesting"). In other words, AND-combining search queries usually
reduces your number of findings.
In the results page, this search is then shown like this:
In other words,
several filters shown in a row are AND-combined
On the other hand, when you search for
crispr OR cas13...
...you are searching for
documents that contain either crispr or cas13. OR-combining search queries usually
increases your number of findings.
And in the results page, OR-combined search terms are shown in one filter, like this:
If now you want to add 'cas12' as another OR-term ("show me documents containing either crispr or cas13 or cas12"), you can click on the filter and add cas12 like this (again, don't forget to hit 'return' when you are done):
Filter boxes are limited in width (imagine how long a filter box would be for a search query consisting of 20 terms combined by logical OR…). So if your query terms do not fit comfortably within the width limits of the filter box, clicking on the filter box opens a popup:
In the popup, you can add further search terms or delete existing ones. In the example above, we added 'tracrrna' (cf. red arrow) as a further search term. After clicking 'OK', all four search terms in the box will be combined by logical OR.
If now you want to search for 'crispr OR cas13 OR cas12...' AND 'machine learning', for example, simply add 'machine learning' in the search field and hit 'return':
The result is:
You can see that
OR-combined terms are in one filter box; AND-combined terms are in consecutive filter boxes.
Logical NOT excludes findings that contain your search term. You can search for
'machine learning' but not 'ecommerce', for example,
by using the '-' operator:
The result looks like this:
Red filter boxes are NOT queries.
How to best build up a longer NOT search query
Sometimes when you do a search, you might want to successively exclude various topics. This is particularly relevant, of course, if your initial search query is very broad, and you now want to successively "slice away" findings that are not relevant to you.
But if you add one NOT query after another, independently of each other, you might get a very long search query that becomes hard to handle. For example, say you searched for "deep learning". This is extremely broad, and let's say you now want to exclude the following topics (plus perhaps some others):
automation, ecommerce, shopping, advertising, finance, recruiting, marketing
If you exclude them independently of each other, you will get a cumbersome query like this:
So here is a better way: After the first query, "-automat*", simply click on that term and successively add the other terms by OR:
Once your string is so long that it doesn't conveniently fit into one query field, a popup window will show up:
There, you can then simply add further queries as new rows. The popup window also makes it easier to bulk-add or -remove terms to or from your query.
And the resulting query is a lot less cumbersome than the one above, where we entered each query independently of each other:
Mergeflow supports the '*' wildcard. Searching for sensor*...
...matches documents containing “sensor”, “sensors”, “sensing”, etc.
biotic and also
probiotic, and any other word that ends with 'biotic'.
Besides AND, OR, and NOT searches, mergeflow also features a NEAR search. For example, a search for…
…only finds documents that contain “internet of things” and “analytics” with a maximum of five intervening words (NEAR2 would allow max. two intervening words, NEAR10 ten words, etc.). Sequence of the search terms plays no role here.
You can also combine NEAR searches by logical “OR”, for instance…
...finds documents that (1) contain “IoT” max. five words apart from “analytics”, or (2) “smart grid” max. three words apart from “analytics”.
How to see Team topics
If you are part of a
team in Mergeflow
, each team member can see the topics that the other team members are following. These topics are suggested when your query matches any part of any team member's topic. Or you can access the list of team topics directly, by typing "team topics" into the search bar:
How to see Emerging Tech topics
Mergeflow has a set of pre-configured searches for
topics built in. You can browse through them by clicking on "Explore emerging tech..." above the search bar:
Alternatively, you can enter "emerging tech" into the search bar. You can also narrow this down to a particular category. For example, if you want to see emerging tech topics from "Materials Science", you can type "emerging tech (mat":
This works in a similar way for all the other topic categories.