Search Engines and Databases

Google is the most commonly known search tool

Google is possibly the most ordinary tool for starting information retrieval. However, one must be critical of the quality of information found through it. Google is capable of finding many kinds of material, many of which do not qualify as sources for scientific work. However, searching from Google may provide you with a general idea of the topic and help you move forward.

Google in information retrieval

“Regular” Google is proficient in finding:

Quick answers for specific factual queries
Information on government, ministries, etc.
Home pages of companies and organisations
Blogs and forums on your topic
News (that may lead you to an original information source)
Free open access publications, e.g., open access articles (Google is useful for checking whether an article or a report you’re looking for is readily available on the internet)

Sorting the results by relevance (results that best match the search terms are displayed first) is often helpful, although sometimes useful information is moved further down the list.

Many factors have their impact on the order of Google search results. For example, the location and frequency of search words on a web page, as well as the user’s own search and browsing history can alter the order. It is also common for advertisements and publications with numerous citations, views, linkages and clicks to appear first in the results.

Advanced search in Google

Google also has an advanced search (available from the settings).

For instance, reports produced by organisations are often available in the PDF format. If you want to target your search on these files, you may select “Adobe Acrobat PDF (.pdf)” from the “file type” menu.

You can also think about which organisation could have produced the information you need and specify that organisation’s domain name in the “site or domain” field.

Google Scholar focuses on scientific publications

When you are looking for scientific sources, switch to Google Scholar. It does not look for regular websites, but instead focuses on reports, dissertations, theses, articles, books, conference proceedings, etc. by scientific publishers, universities and other organisations that are considered to be scientific. Google finds “grey literature” as well, but the amount of content it covers is not known.

However, not all content found by Google Scholar qualifies as scientific. Many theses from universities and universities of applied science’s appear in the results when searching in Finnish, in particular. Thus, searching in English is often more rewarding.

Google Scholar is useful for:

Starting the research. You may find some relevant articles to help in the beginning, but don’t place your trust in them alone.
Finding the “key factors”, the most referenced authors, for your topic.
Brainstorming terminology and search terms.
Finding additional sources via related articles and the number of citations.

Google understands clauses (questions, inflections, articles, prepositions, etc.) quite well, and automatically searches for terms related to the topic. There is significance to the order of the search terms in Google, and the most important terms should be typed in first.

It is possible to use advanced search, AND and OR operators, as well as moderately functional phrase search on Google. Boolean operators must be written in capital letters.

Note! What works in Google may not work in databases.

Click here to view the accessible version of this interactive content.

In case you can not access the full-text, log in to UEF Primo and check the availability through there. A UEF account may allow you to cross the paywall.

Databases for scientific information

Research information is sought from databases, which are constantly updated collections of scientific publications in a certain field (narrow or broad). These are often called bibliographic, reference, and or library databases.

The more thorough and systematic your search needs to be, the more evident it is that you need to search specifically from the field’s (article, reference) database. For example, for systematic overviews the information retrieval must always include multiple article databases.

Databases feature better limiting functions than the different versions of Google. In databases it is also possible to confirm the information on the sources taken to the database, as well as the criteria for their selection: databases often include listings of the scientific journals whose articles are in the database, and the criteria for selecting publications may be disciplinary-specific.

UEF Primo: collection database and a portal to the electronic materials of UEF

UEF Primo is the search service for the printed and electronic materials of the University of Eastern Finland. In UEF Primo, you can renew your loans, make reservations, and save your own favourites. UEF Primo makes it possible to display all the library’s electronic material: bibliographic databases, electronic journals, e-books, and other online resources. In addition to licensed online resources, openly accessible material from the web is also featured.

UEF Primo includes both domestic and international material. Resources like books should also be searched with English search terms, as it enables e-books to appear in the results more frequently. On the other hand, Finnish search terms may also retrieve material in English and other languages, as printed books, in particular, often feature YSO index terms. More about YSO can be found in the Index Terms and Glossaries -section.

You may search international articles with UEF Primo by focusing the search from the drop-down menu: international articles, or filtering from the right-side row of the results list: resource type. Domestic journal articles should be searched from the national Finna (see below).

Learn how to use UEF Primo with UEF Primo Search Guide.

Click here to view the accessible version of this interactive content.

UEF Primo can help you to find databases. Databases can be found by typing their name or its part into Primo’s search field and accessing the database’s website from the link in Primo’s record. Be sure to log in to UEF Primo to get wider access!

Finna: data on the collection of Finnish libraries and domestic articles

The national Finna contains information on all the resources included in Finnish library collections, with information on electronic resources included.

It is recommended to search from Finna when you wish to comprehensively cover Finnish book resources (as well as studies, dissertations etc.), or when you wish to confirm whether a work, book or a journal is available in Finland.

Information from domestic journal articles is also stored in Finna. You may access this information by limiting the content type to “journal/article” from menu. However, professional, and popularised journal articles will also appear in the results, and they can not be excluded by criteria of science or peer-review. In these cases, it is possible to check the Publication channel search rating granted to them ensure that they are scientific.

See Finna’s own instructions.

International databases widen information retrieval

International reference databases can provide you a grasp on what has been written about your topic.

For example, articles in international scientific journals are important research literature. Databases often contain information on book resources and conference proceedings in addition to journal articles. Articles are either available in full-text or merely as citations depending on the database.

Examples of multidisciplinary databases:

Scopus
EBSCOhost Academic Search Premier
Web of Science

In addition to multidisciplinary databases, several disciplines have their own, discipline specific databases. These databases can be viewed through the UEF Library’s information retrieval and training -website.

Also, check out our practical examples of using databases on the Search Examples page.

Artificial Intelligence in Information Retrieval

Artificial intelligence (AI) has recently made its way into public discourse through easy-to-use chat bots such as ChatGPT and Microsoft Copilot. These conversational AIs are based on generative language models. Language models are built on great masses of text (Large Language Model, LLM).

General topics in short

The bot that uses generative artificial intelligence is used by typing in questions in plain, natural language. The AI uses a language model to generate answers from text based on how words most likely follow each other in natural language. The AI works interactively, allowing you to steer its answers in the desired direction by asking new questions.

The AI also learns from its previous answers. The answers often seem reasonable and are grammatically correct, but there is often no guarantee that they are based on factual evidence. The AI tool has no semantic understanding of the content of information. AI can also hallucinate, which may result in answers that are easily identified as absurd.

When using generative artificial intelligence, you need to pay attention to data protection and keep your private information safe, so please read to terms of use carefully. Read also UEF guidelines: The use of AI in teaching and research. Many applications may also require a personal user account.

How can conversational AI help?

Generative AI is not a database for directly providing verified facts or relevant literature references.

It can, however, offer a helping hand for an information seeker:

Use AI for brainstorming search topic
Make your search topic and search questions more clearer or wider. AI can be particularly helpful in analysing a new and unfamiliar topic.
Let AI suggest search terms
AI assists in the analysis of a search topic and ideation of search terms and search queries. It can also help translating those terms, for example from Finnish to English.

You should not rely blindly on search terms suggested by the AI, but instead verify them using other sources. The provided logic of search queries, i.e., the use of Boolean operators and parentheses, may also leave something to be desired.

The easiest option for testing the AI’s suggestions is to search from a scientific database in your field and see if the results make sense. You can then fine-tune the search terms and phrases based on the results and your own knowledge.

Example questions

Can you suggest keywords for this topic:
Can you create a database search query with keywords and Boolean operators for this topic:

Some examples of applications

Microsoft Copilot
Open AI ChatGPT
Google Gemini
For finding subject terms:
- National Library of Finland Finto AI
- National Library of Medicine MeSH on Demand

Aritificial intelligence in scientific databases

In addition to chat bots, artificial intelligence has also been used to develop bibliographic databases, such as Elicit, Consensus and Scite, which contain peer-reviewed information.

In Elicit, Consensus and Scite, search terms should be given in natural language. After the search, the AI sorts and groups the publications in order to find commonalities. This is intended to make the further work of the information seeker more efficient by more or less eliminate the time-consuming process of browsing through the search results.

The main problem with these databases is the lack of opacity of the search process. The result sets of the searches are usually large, and there is no way of checking how the sorting works and whether relevant material may be excluded from the results.

The well-known scientific databases, Scopus and Web of Science, are also looking to introduce some form of AI in the near future. The explicit aim is to help information seekers to quickly find summaries of new as well as lesser-known research topics.