7 reasons why you should try Lens.org (updated to version Release 5.16.0 — March 2019)

Lens is in my book one of the most interesting Scholarly discovery/ citation index tools to have emerged in 2018. I am not saying this lightly as 2018 was the year crowded with new discovery services like Dimensions, 1Findr (now acquired by Elsevier), Meta (in closed beta at time of writing) and more.

Owned by the non-profit Cambia, it promises to be free of charge for all (no freemium model) and further more claims to safeguard your privacy with no use of Google Analytics or other cloud based click-trackers.

Of course all this isn’t worth anything if the tool isn’t useful. Given the dominance of Google Scholar as a discovery tool, there seemingly isn’t much room for another discovery tool. But Lens I think is more than just a simple discovery tool, it actually allows you to explore and analyze the data in ways Google Scholar is unable to match, thanks to a blend of powerful filters, facets , customizable visualization capability and bulk export functions.

In many ways it’s a substitute for citation index tools like Scopus and Web of Science and in some ways go beyond it. In particularly it has many features (e.g. easy bulk export of 50k records, visualization tools) that make it suitable if you want to analyse output at institutional scale except it’s totally free.

1. Lens has one of the largest Scholarly record index out there

Size matters and Lens has over 197 million Scholar records sourced from Microsoft Academic (the major source), Pubmed , and Crossref. This is as big as it gets, and only Google Scholar is probably bigger. (Compare Scopus only has 60 million records). It also merges citation counts from Microsoft Academic and Crossref. Given that various studies have shown Microsoft Academic has probably one of the biggest citation indexes second only to Google Scholar, Lens is obviously building on a strong foundation.

But why not just use Microsoft Academic?

2. Lens also includes one of the most complete patent indexes out there

But did you know Lens actually began life as a patent search tool more than a decade ago? Lens has over 111 million patent records from over 95 jurisdictions including 58M patent families, 550k biological patents and 295 patent sequences.

Tools like Lens Patcite, allow you to either explore cited works found in patents (allowing you see if your scholarly works have been cited in patents) or conversely enter patents to see what Scholarly works have been cited.

In fact Lens is well recognised as a leading source of patent citations and information and even the likes of Microsoft Academic have announced a partnership with Lens.org to improve their patent coverage and the patent cites from Lens feeds into the Crossref Events data API

The advantage of Lens is that it incorporates both Scholarly and Patent literature so you can not just look at the Scholarly record side but you can also easily see the impact of Scholarly works on patents.

Patents that cite your set of Scholarly record

You can see the number of cites from patents to your scholarly work sets and with a click you can filter down to Scholarly articles that have been cited in patents. I have done some informal tests with Scopus/Scival and the numbers found by Lens is much higher….

Using the cool visualization features (more on that) you can even see the applicants that cite your Scholarly work for a set of Scholar records.

Applicants who have cited works from my institution (Singapore Management University)

3. Lens has a powerful advanced search allowing you to filter by an amazing array of fields like author affiliation, funding info, MESH (and other Pubmed ), open access availability and licenses (via unpaywall) and more

Powerful structured Scholarly search allows searching and filtering with dozens of field options

Lens is more than just a literature discovery tool. Sure Google Scholar may be the tool of choice for many, but can you slice and dice the results by any of these fields?

-Lens Scholarly ID
-citation identifiers
-publication date
-publication type
-authors (first and last name, order, affiliation)
-start end pages, volume, issue
-funding/grant information
-keywords (PubMed only)
-mesh_term (PubMed only)
-Field of Study
-chemicals (PubMed only)
-clinical_trial data (PubMed only)
-citing patents
-scholarly citations
-recommended works
-references (string with identifiers if available)

The combination of data from various free sources like Crossref ,PubMed, Unpaywall, ORCID ,CORE and more means you can do all sorts of advanced exploration using filters and facets.

Advanced facets are available in Lens on top of advanced search

In particular it is the only free tool(besides Microsoft Academic from which it draws affiliation data) that can allow you to filter by affiliation, which is very helpful for librarians and administrators who want to study research output at an institutional level

For instance I was recently wondering how many of our articles were published with funding and were open access. With Lens you can do it with a few clicks and facetting

Papers with Funder info and the % OA per Funder

4. Lens allow you to create powerful flexible visualizations by type of field or by chart type

Prior to the March 2019 version, Lens provided fixed visualizations. It now provides pretty powerful flexible visualizations you can customize.

As usual run a search and get a set of results you are interested in . In my case, I will run it over papers written by authors of my institution (n=6,794 records.). There are some default set of visualizations you can use but you can choose to create your own.

Choose the item you would like to analysis

First choose if you want to run an analysis over Scholarly Works, Citing Patents, Institutions, Authors , Funders , Journals , Field of Study, Countries, Open Access, Substances or click custom (more on that later).

Say you click on Open Access, and you get some suggested visualizations

Choosing visualizations around Open Access

Say you select “Journals/Source Titles by Open Access Status” you get the following visualization.

Visualization of proportion of OA per publisher

Think that is all? You can further customize this visualization by clicking on the settings dropdown and change things like color, number of publishers to be listed , to convert it to raw numbers, change from stacked to grouped etc.

Changing settings in the visualization

Most importantly by selecting different facets you can change to use pretty much any field Lens has to do different visualizations.

Lens rovides quite a few main types of visualizations including Horizontal Bar Chart, Vertical Bar Chart, Grouped Bar Chart, Stacked bar chart, Line chart, Scatter plot and more. You can see them by selecting “Custom” in step one of the visualization process.

It can get quite flexible below shows one of my favourites — Top funders per institution

These customization features put it comfortably above expensive commerically licensed databases like Scopus or Web of Science. To be fair these 2 databases have add-ons in Scival and Incites that provide such visualization capabilities but those are even more pricey.

That said as I write this while Lens provides scatter plots for comparing top 1000 papers with Scholarly citations or patent citations there is a surprising obvious gap. There is no way for example to compare the average citation count (or any similar measure) across say 2 groups.

5. Lens allows you to easily create collections, saved searches , Visualizations and batch export up to 50k records.

Lens doesn’t just have a powerful search, there are other useful features like the option of creating saved searches (with email alerts), saved collection(closed, limited,restricted settings), notes and more. The custom visualizations you created earlier? They too can be saved in a dashboard.

My saved collections

Want me to blow your mind further? You can export up to 50k records per batch! This is more than what you can do for expensive commerical tools like Scopus or Web of Science

To be fair not all fields can be exported but most of the ones I’m interested with (eg references, funding info etc) generally can be.

6. Easy to use interface

The interface of Lens might not be as slick looking as some databases, but it is very serviceable.

Take the process of add all records for all your results (across all pages) into a folder for exporting. I confess when I try to do this in Scopus and Web of Science, I’m occasionally stumped particularly if I haven’t used it for a while. With Lens, it is as straight forward as you can imagine , just click on the checkbox, click on “All results” in the popup and then “Create collection”

There a couple of things I would tweak e.g. allowing one to modified advanced structured search rather than starting from the scratch but it’s already excellent.

7. Lens links to free and paid copies

Lens provides two options to help you access a copy of the paper. First there is a link to open access copies using Unpaywall data. Secondly if you have institututional access and your library has set it up, you can click on the “Find full-text at your institution” option.


I have just scratched the surface of the capabilities of Lens and the tool is developing very quickly.

Given that Lens is free , there seems no reason not to give it a try. Particularly if you have no access to expensive tools like Scopus or Web of Science. Even if you do have access like me, give it a try and compare you might be surprised by how well Lens holds it owns even against expensive tools.

The main concern I have with Lens is the reliability of the data which at this point is still not well established. An API would also be very nice.

That said, I have done a mini study to do a preliminary baseline

  1. Export all papers from Scopus tagged as belonging to my institution
  2. Export all papers from Lens tagged as belonging to my institution
  3. Match by doi
  4. Run a correlation
Scopus vs Lens comparison (R=0.96)

I guess nobody is surprised the correlation is nearly perfect at 0.96 (statistically significant of course)

Still there is a lot more of work to be done to establish the reliability of this statistic.