My visit to AgroKnow and the Author disambiguation project in AGRIS

I am a metadata librarian from the Chinese Academy of Sciences - which is one of the AGRIS strategic partners. Currently l am visiting the FAO AIMS team and working with the AGRIS and my contribution is the author disambiguation project. Recently l visited one of the AGRIS partner institutions AgroKnow, in Greece. Below l re-share part of the blog published on their website about my visit. 

What exactly is the problem of author disambiguation?

As indicated by analytics of FAO AGRIS, in many cases people are searching for publications using authors’ name. This calls for a search engine that will be able to understand the variations of an author’s name but also to distinguish between authors with similar names. For instance if a person like Giannis Stoitsis was also using in the past the name John Stoitsis in publications then the system should be able to understand that this is the same person by using extra information and a robust algorithm. This problem is known as authors ambiguity.

The solution to this problem is easier when we have rich information for the author. Such information can be apart from the first name and surname, the location, the area of interest, communication details, affiliation. But in the case of bibliographic references the information mainly includes the full name, the affiliation and only in some cases the location. This means that we need to link to external sources to get more information to solve the problem. This is exactly the point in which linked data can help. Some years ago the most important organisations in the research sector agreed on developing ORCID, an open, non-profit, community-driven effort to create and maintain a registry of unique researcher identifiers and a transparent method of linking research activities and outputs to these identifiers. ORCID provides an URI for each researchers and additional information? The most important is that it provides also the property also known as for a person somin the case of this strange person called Stoitsis, ORCID keeps both Giannis and John as first name. The ORCID page for Giannis Stoitsis is

In FAO AGRIS, that is powered by AKStem, apart from the basic information (that we collect in each metadata record of a bibliographic reference) we will use also AGROVOC terms which is an indication of the area of expertise of a researcher and the list of co-authors. We will identify person entities using the FREME NER service which provides for identified entities also the ORCID URI if it is available. An example of using FREME NER for my self can be tried here.

What’s next? We will validate the designed process with a sample of AGRIS authors DB. Of course, we will share the results. Stay tuned.

Add comment

Log in or register to post comments