Patents

Solution

Last Updated (Tuesday, 18 November 2008 16:23)

3) How SOPHIA addresses these problems

SOPHIA’s patent searching technology can address all the limitations outlined in the challenges of patent searching section because it searches based on the meaning of what you are looking for not simply on the presence or absence of the key terms you enter. It can do this because it uncovers the implicit hidden relationships between patents, based on their content. This ensures that if patent A is of interest SOPHIA understands Patent B will also be relevant because it is very similar based on its meaning to Patent A. Patents B could contain terms that are synonyms to those in Patent A and therefore it could remain hidden to conventional Boolean search techniques. Not so with SOPHIA.

With SOPHIA you no longer need to craft complex Boolean queries but instead can present a paragraph of natural language text outlining your ideas – this is much more intuitive and much more efficient and effective. By presenting a paragraph description (i.e. an abstract) SOPHIA will identify the key themes within the patent corpus that are most relevant to the query and identify from each theme the actual patents that are most semantically relevant.

Invented legal terms are not a critical issue as SOPHIA focuses on the overall meaning of the patent not just specific key words. If portions of the patent are missing the impact of this will be dampened for similar reasons.

How SOPHIA works

SOPHIA automatically structures patents into thematic groups based on their content.

All patents associated with a particular theme are semantically related.

SOPHIA discovers the implicit semantic links between patents – so it can determine which patents are most relevant to your needs.

When presented with a query SOPHIA identifies the most relevant themes and then within each theme, the most relevant patents to your needs.

Benefits of themes

You automatically discover information you didn't know existed

You understand the different contexts that exist for you query

It ensures you interpret information correctly – for example, Ancient Egyptian texts can only be understood properly when there are a number of texts together. To understand the meaning of 1 in isolation is impossible – This is called intertextuality. It is the same with patent information, in order to understand the significance of one patent it needs to be understood in the context of all other patents in a similar semantic space.

It facilitates being able to link patents based on meaning

SOPHIA does away with the need to form complex Boolean queries