Making Search Meaningful The Pitfalls of Multi-Meaning Keywords
At their core, search engines are internet encyclopedias that allow users to find and filter out relevant information. With any given search engine, it takes users some skill to find exactly what they are looking for. Users must understand how a search engine works and how search queries are interpreted.
Adding to the complexity of today’s internet search techniques is the problem of multi-meaning keywords—words or phrases that have several meanings depending on how they are used in context. Imagine searching on a term such as “sewer,” “copper,” “gloss,” or “ship” and expecting to find a part or product to solve a manufacturing need. Instead, search engines may return a variety of consumer-related content that is unrelated to a specific business topic.
As a further example, Merriam-Webster defines the word “polish” in a myriad of ways:
- to make something smooth and shiny by rubbing it
- to improve something or make something better than it was before
- to put nail polish on (such as a fingernail or toenail)
Words or phrases that can have two or more meanings are referred to as “polysemous” and can pose challenges for nearly anyone involved in search engine use:
- Search engine developers: attempting to identify the true intent behind user searches
- Search engine users: sifting through search results unrelated to what they are really trying to find
- Website owners: finding their pages in search results surrounded by sites offering something very different from what they offer
- Online advertisers: bidding on certain words or phrases as sponsored results for searchers who may have no interest in these ads
If a user enters the word “polish” into a search engine, the results they see will likely be filled with pages related to Polish heritage, language, and culture; fingernail polish and nail salons; the Wikipedia entry on Polish (anything from or related to Poland), and other consumer-related content. This occurrence would undoubtedly frustrate users looking to connect with manufacturers of polishing machines and related products.
How do search engine developers go about solving the problem of multi-meaning keywords?
Potential Solution #1: Displaying Query Refinement Suggestions
One solution might be for a search engine to show query suggestions along with search results—adding more words to the original query based upon previous queries from other searchers.
A search engine could look at its query logs to determine how often searchers made changes to their queries and associate these follow-up queries with the originals. Many searchers, faced with a mix of results like those for “polish,” might change their original query to include an additional word or phrase that makes it more likely that search results will be relevant to what they intended to find.
For example, someone searching for information about the act of polishing might add a word like “equipment” to their search. Search results would then include pages about deburring, grinding, honing, and related polishing processes. If enough users refine their searches in this way, a search engine might start showing “polishing equipment” as a link within search results for a search for “polish.”
However, given this approach, search engine developers would have to continuously collect enough data to decide which additional query refinements to display—a very time consuming endeavor.
Potential Solution #2: Using Related Words to Understand Meaning
Another approach could be for search engines to review web documents that include polysemous words and then identify the other words that appear nearby. Thus, a search engine could locate words or phrases that might be related to the various meanings of the original polysemous words.
For example, one meaning of the word “sewer” is a system of pipes used to transport sewage. Using this definition, it may be possible for a search engine to display applicable results if it appears near words such as “sanitation,” “drain,” or “sewage.”
Being able to distinguish between the different contexts of a word that has more than one meaning is a significant undertaking because a search for sewer, a drainage system, is different than one for sewer, a person who sews clothing and other items.
A patent application filed by Yahoo in 2007 and published in 2009 describes a process that search engines could use to determine if a word or phrase has more than one meaning based on the words that tend to co-occur in web documents including that word.
Patent Abstract: Method and Apparatus for Discovering and Classifying Polysemous Word Instances in Web Documents[A method and apparatus for discovering polysemous words and classifying polysemous words found in web documents. All document corpi in any natural language have words that have multiple usage contexts or words that have multiple meanings.
Semantic analysis is not feasible for classifying all word occurrences in all documents on the web, which contain trillions of words in total. In addition, semantic analysis typically cannot distinguish multiple usages of a given meaning of a given word.]
Content Author Solution: Distributing Web Content for Usability
The Yahoo patent filing provides details on how search engines might count how frequently or infrequently words appear next to each other in web documents. The patent also describes a process of breaking down web documents into smaller blocks of words.
One reason for web authors to adopt the process of “chunking” content is that some pages may be very long, and targeting words found on those long pages may result in creating relationships between words that are not necessarily related. Another advantage is that a content page may contain multiple postings that are not related in primary subject—providing search results that are not specifically targeted for a search term.
According to some estimates, more than 40% of English words have more than one meaning. This phenomenon poses clear challenges for anyone involved in search engine development and usage.
Search engine users, web content publishers, and online advertisers should keep polysemous words in mind as they perform searches or create content intended for optimal search engine results. Specifically for manufacturers, understanding the effects of polysemous keywords enhance advertising efforts and offer competitive differentiation.Review All Articles