4.1.2 Search Theory

Date Published

Search theory has developed over the last century as a statistical analysis that provides a rigorous approach to searching for lost objects. Spearheaded by the work of coast guard and search & rescue organisations, much of the theory can be applied to searching for prior art documents.

The key principle of search theory is that the search is not a linear process, but instead is a dynamic iterative exploration of the search field.

The theory can be distilled down to the following four concepts:

i. identify the areas most likely to provide results,

ii. carry out focussed searches in that area to look for the item,

iii. reflect on how successful the search was and identify what can be learnt,

iv. if required, move to the next area to search based on the learning from steps (i-iii).

From a patents perspective we call this an “iterative search” using your own knowledge of the prior art and the input of your 3PT team to work out what you are trying to find and where it is likely to be located. The search should target that area first using controlled language, classification symbols and/or keywords that have the highest likelihood of success.

Searching precisely allows you to assess the effectiveness of particular classification symbols and search terms. This approach should also exclude irrelevant art meaning your search will be more effective.

For example, the documents that you view might suggest additional classification symbols, keywords or search strategies. Equally, a particular search term might not be as relevant as you originally expected; or a particular classification symbol may not be precise enough. Adopting an iterative approach whereby you reconvene the 3PT, or even run the search with them, provides an opportunity to reflect on what aspects worked well and what didn’t.

If you could not find relevant art, or you believe there is more relevant art elsewhere, then you can iterate your search by moving to the next search area in a much more informed manner.

Subsequent searches might still be targeted, for example if the technology you are looking for is spread across a range of classification symbols. Equally, subsequent searches might embody the same search terms from earlier searches, but the search starts more broadly. Later searches may also explore different approaches for example using a full text database instead of an abstract database.

In summary, searching is not a prescriptive process. A series of informed searches (iterative searches), using a combination of different search techniques and focusing on different combinations of features carried out in an iterative manner is likely to maximise the probability of identifying the most relevant art.

Iterative searching involves:

  • Constant evaluation of retrieved documentation which informs the initial strategy and develops subsequent search iterations.
  • The initial focus of the search is continuously revised and adapted based on intermediate search results and consultation with the 3PT.
  • Search strategies are not fixed but adapt to the knowledge gained during the search.

The following diagram summarizes the iterative searching process: