Artificial Intelligence and Citation Patterns

Increasing use of technology in the production of scientific knowledge is of course an uncertain path, because even the present is unpredictable. I am not an academic and have no citations to my name, however I have been discussing citation patterns with an Associate Professor at my university. We thought about how this could be influenced by machine learning techniques or the field of artificial intelligence in general. Lo and behold he sent me a picture from one of the groups he was participating in that led me to think it was worth looking closer.

The picture was the following:

Inter-field citations concentration

As you can see there is a highest concentration towards the top five percent with the same papers being cited over and over. It does seem that perhaps the field of academia is experiencing a social media effect. The echo bubbles that were ‘smaller’ previously may now be present in the academia of the present (although this is a large and bombastic generalisation). There can be something to be said for the increase of citations between fields too, perhaps this is a positive move? Is it easier now to find information from other fields than your own thanks to Google Scholar?

Google Scholar can do a lot and in a way it is a social media platform in the way that there is user profiles and that different authors can spot the citations that have been given to the various papers they have published.

“It has been shown it’s possible to spoof Scholar, with fake articles of randomly-generated words accepted, and even a fake author who became the world’s 21st most-highly-cited scientist.” — Louis Coiffait in Social Science Space

I could not easily locate how many searches that occur on Google Scholar and would love to find more stats. The East Carolina University posted a few suggestions for and against Google Scholar. So first the advantages:

1. Google Scholar is familiar and relatively simple to use, much like Google.
2. Google Scholar allows users to search for a find a wide variety of materials including articles, books, “grey literature” like conference proceedings on a vast number of topics.
3. Google Scholar allows for you to see articles related to the one that might interest you, how many times an article has been cited and by whom, and provides citations for articles in a number of styles.
4. Google Scholar can display links to articles and books held through ECU Libraries. For more information, see Using Google Scholar from Home tab.
5. Google Scholar allows you to save both citations and articles to read later.

Then the less disadvantages:

1. Google Scholar’s coverage is is wide-ranging but not comprehensive. It can be a research source, but should not be the only source you use.
2. Google Scholar does not provide the criteria for what makes its results “scholarly”. Results are often vary in quality and it is up to the researcher to determine which of the results are suitable for their purposes.
3. Google Scholar does not allow users to limit results to either peer reviewed or full text materials or by discipline.
4. Google Scholar does not provide notice of when its materials are updated.
5. Google Scholar’s citation tracker can be difficult to use and inaccurate.

I do want to look more into this aspect going forward. I have read that data is available from Google Scholar in some way, shape or form, but I am yet to find any data. However the general questions do not answer mine (is it not general?).

I am currently not one to judge what influence Google Scholar has on citation patterns, however I think it is an interesting aspect of contemporary research to be aware of and conscious about.

I hope you enjoyed this short article. This is day 134 of #500daysofAI. I write one new article about or related to artificial intelligence every day for 500 days.