Search for Terms and Phrases (Full-Text Boolean Searching)

 Connectors Summary:

This case study accompanies the above video and explores how each component of the search strategy summarized above affects a set of search results in real time.  The numbers expressed below are taken from the above video

Step 1: Searching by Category Only (No Search Terms)

Objective: Find Financial Statements and Management Discussion and Analysis (MD&A) documents from Consumer Products and Industrial Products companies.

Steps:

1. In the SEDAR Filings dataset, locate the Industry and Document Category criteria among your usable criteria to the left of the Search button

2. If you don't have Industry and Document Category criteria visible (please check well - they are most likely visible already), then add them by clicking + Add Criteria in the upper left, underneath the full text search field.

3. Once added, make the following selections:

   1. Under Industry, select Consumer Products and Industrial Products.
   2. Under Document Category, select Financial Statements and Management Discussion and Analysis.
   3. Set Filing Date to Last 2 Years.

4. Click Search.

Results:

• 1000 documents retrieved.

• These are all financial statements and MD&As, but there’s no way to tell if they mention net profit or sales growth without manually searching inside each document.

• This search does not use a full-text component so we have no idea how many of the results contain the language we're interested in.

Step 2: Adding Search Terms with AND

Question: Do I need to find specific words or phrases in these documents?

Steps:

1. Type net profit and sales growth 

2. This instructs Avantis to limit your results to only documents that have these two exact phrases anywhere in them. 

   1. You do not need to enclose the phrases in quotation marks.  Any words without recognized connectors in between them will be taken as exact phrases

3. Click Search again.

Results:

• 34 documents retrieved (compared to 1000 before).

• This means 966 (96.6%) of the original results were irrelevant noise.

• Each document now highlights where "net profit" and "sales growth" appear

• If you open your left panel and choose the Keywords tab, you will see every context within which the phrases occur in the document. 

• However, the search does not account for variations in phrasing (it is not even looking for plurals) or related phrases.

Takeaway: Even this simple search saves 95% of the time compared to manually checking 1000 documents.

Step 3: Using a Proximity Connector ("w/n")

Questions:

• Are my exact phrases too limiting or too exact?

• Do I need to find related words in the same discussion, even if I can't predict all the various ways they might be used in a phrase? 

Steps:

1. Update your existing search

   1. Old version:  net profit and sales growth 

   2. New version: net profit AND (sales w/5 growth)

      1. AND does not need to be capitalized - this is only done for effect, with the reader in mind  
         1. Avantis reads AND and and as exactly the same word
   3. The w/5 instructs Avantis to retrieve only documents that feature the exact phrase "net profit" in them as well as any occurrence of the word "sales" that appears within 5 words of the word "growth"
   4. Had you used w/10 instead of w/5 Avantis would look for both words within 10 words of each other instead of 5 words
   5. Had you used w/100 instead of w/5 Avantis would look for both words within 100 words of each other instead of 5 words
   6. Proximity connectors allow you to set maximum distances between your terms with absolute control and precision
2. Click Search again.

Results:

• 81 documents retrieved (more than double the amount in Step 2).

• This search finds variations like:

  • "sales growth"

  • "growth in fourth-quarter sales"

  • "growth in international sales"

• The W/5 proximity connector ensures "sales" and "growth" appear within 5 words of each other, making the search more flexible while still being precise.

Takeaway: Proximity searching expands results without adding noise, helping you capture relevant discussions.

Step 4: Adding Synonyms (Using "or")

Question: Can any of my search terms be replaced with other words that would yield good results?

Steps:

1. Update your existing search

   1. Old version:  net profit AND (sales w/5 growth) 

   2. New version: net profit AND (sales or market w/5 growth or increase)

   3. This will include results where the word market is within 5 words of increase or grow

2. Click Search again.

Results:

• 250 documents retrieved (more than 7x the amount in Step 2).

• This could be expanded further with other synonyms like improve, augment, ramp up, double in the place of grow or increase depending on your needs and appetite for variation.

  • to do this you'd simply use "or" in between each synonym - eg - net profit AND (sales or market w/5 grow or increase or augment or double or ramp up)

Takeaway: Using synonyms helps capture different ways the same concept is expressed, significantly improving search coverage without much effort.

Step 5: Adding a Wildcard ("*")

Question: Do I need variations of the forms of my search terms (plurals, suffixes, tenses, etc.)?

Steps:

1. Update your search

   1. Old version:  net profit AND (sales or market w/5 growth or increase)
   2. New version: net profit* AND (sales or market* w/5 grow* or  grew or increas*)
   3. This instructs Avantis to retrieve only documents that feature the exact phrases net profit or net profits in them as well as any occurrence of either the word sales or else any variation of the word market that appears within 5 words of any variation of the word grow or increase
      1. Note that simply putting an asterisk after grow will not get the word grew since it doesn't start with the same four letters as grow so we add grew as a synonym or separate word
2. Click Search again.

Results:

• 380 documents retrieved (almost 11x the amount in Step 2).

• Wildcard (*) expands search terms:

  • profit* → profit, profits (since "profit" is part of a phrase, this won't likely trigger words like "profiting", which are unlikely to follow the word "net"
  • market* → market, markets, marketplace, marketplaces
  • grow* → grow, grows, growing, grown, growth
  • increas* → increase, increases, increased, increasing

• We had to manually add grew as a synonyms since it won't be generated by grow*.

Takeaway: Using wildcards increases flexibility without requiring separate search terms for every variation.

Final Thoughts

Key Takeaways:

✅ Even a basic full-text search (Step 2) eliminates 95% of irrelevant results.
✅ Proximity connectors, synonyms, and wildcards dramatically increase relevant results.
✅ Combining these techniques can make your searches exponentially more effective.

The best way to get into the habit of terms and connectors (Boolean) searching would be to apply one step at a time until you are confident enough that you can mix the steps together as you type out your search.  There is nothing wrong with building your search out step by step every time you want to run a comprehensive search.  

Would you like to see an interactive demo of these techniques? Let us know!

CASE STUDY, CONCLUSIONS:

1. Even the most basic search, the search in step 2 above, cuts out 95% of the noise you would have to wade through on SEDAR or any method relying on Ctrl+F to find words in documents
2. The most specific search,  the search in step  5 above, using (1) a proximity connector , (2) synonyms(separated by an or ), and (3) wildcards, found over 11 times as many relevant documents as the basic search,  the search in step  2 above, did
3. By sorting results by rank you will start with the most relevant results (those that have the search terms most frequently occurring and most tightly clustered together) first so you don’t have to look all of the 380 results but can just look at the highest/best matches among them
4. Open the viewing panel in the left margin of the document and click on the keywords tab to see all the uses of your search terms, throughout the document

To search with connectors

Video #2:  Basic Terms and Connectors Searching