HathiTrust Research Center#
-
The HathiTrust Research Center (HTRC) enables computational analysis of the HathiTrust corpus.
Log in with your partner institution (SMU) account to access the largest number of volumes and features.
Click here for a list of HTRC tutorials to walk you through the steps for using HTRC tools and data.
Worksets#
Recommended Worksets are curated worksets created by real researchers of projects that utilize resources from the HathiTrust Digital Library and can be analyzed with tools built by the HTRC.
Algorithms#
HTRC Algorithms are web-based, click-and-run tools to perform computational text analysis on volumes in the HathiTrust Digital Library.
In a basic text analysis workflow, a researcher:
Gathers digitized text (text that has been scanned and OCR-ed) Note: OCR (Optical character recognition) refers to the mechanical or electronic conversion of images of typed, handwritten or printed text into machine-encoded text.
Applies computational methods to that text, such as word counts, classification techniques, and topic modeling
And then analyzes the results generated by the algorithm or technique
HathiTrust Research Center (HTRC) enters the workflow at the points of providing digitized text at scale from HathiTrust Digital Library (HTDL) and providing tools and services that enable computational research.
The researcher, of course, still brings her own analysis to bear on the results.
HTRC tools#
Use the HTRC algorithms as an example of how to use a pre-built tool for text analysis. The HTRC provides some algorithms through HTRC Analytics, and they can be used to analyze HTRC worksets.
An algorithm is just a way of saying a computer function - text goes in, process happens, and results come out.
HTRC algorithms can extract, refine, analyze, and visualize worksets. They can basically perform “plug-and-play” text analysis.
Because they are built into HTRC Analytics, they are mostly limited in how much they can be tweaked or customized. The algorithms are primarily for users who don’t know how or don’t want to work with custom code. It can be a good tool for learning and just trying things out.
How do you choose which HTRC algorithm to use? Naturally, it depends on what you want to do.
Some of the algorithms are task oriented. For example, there are ones for running spellcheck reports, for creating word counts, for visualizing more frequently used words, and for generating a list of named entities.
Others algorithms are more analytic, like one that generates topic models, and another that compares salient words in two worksets.
Try to gain some basic understanding of your tool. Reading some relevant documentation is especially important when the tool is within a “black box”. Speaking of understanding your tools, it’s important to note that the HTRC algorithm only has two parameters you can set right now, so it’s not suitable for really robust topic modeling. But for teaching and exploration of HT text specifically, the HTRC topic modeling algorithm can be a good place to start.
Login to HTDL & HTRC#
First, go to the HathiTrust Digital Library (HTDL) interface.#
Click on the “LOG IN” button on the right to sign in.
If you are affiliated to an HT partner institution (such as SMU), select your partner institution from the list and click on continue, then follow the directions for institutional log in.
Once you are logged in, click on the “FULL-TEXT” tab to search in full text. Now you can start thinking about a search query to find volumes for your collection. Try building your own collection.
View search results. You can facet in the sidebar, then select volumes.
When you’re satisfied with your selection, click on the “Select Collection” bar and choose “[CREATE NEW COLLECTION]” from the drop-down menu. Then hit the “Add Selected” button on the right.
A pop-up window will appear that prompts you to add some metadata to your collection before the system creates it for you.
Fill in the name and description of your new collection.
You can choose to make the collection public or private, and we recommend writing a short description whenever you make collections public.
When done, click on “Save Changes” to create your collection.
After the collection is successfully created, you should see a a confirmation message above the search results.
To view your collection, click on “My Collections” near the top right of the page.
This will bring you to all your collections. *You can manage your collections here by viewing collections, changing the public or private status of a collection, and deleting collections you don’t need.
Click on the title of the collection you just created to view it.
While viewing a collection, you will see Collection Tools
There will be an option for a sharable link to this collection (only sharable and usable in HTRC if you choose to make your collections public)
You will be able to see the title and description of your collection, as well as all the items in it.
In order to import your newly created collection into HTRC Analytics for analysis, we need to download the metadata of the collection first.
Click on the “Download Metadata” button on the left sidebar and select the “Download Item Metadata: Tab-Delimited Text (TSV)” option.
You will not be able to upload the JSON metadata of your collection to HTRC Analytics, so make sure to select the TSV option.
Second, go to the HathiTrust Research Center (HTRC) Analytics interface.#
Go to HTRC Analytics
Look for the sign in/ sign up section on the top right part of the page.
If you are affiliated to an HT partner institution (such as SMU), select your partner institution from the list and click on continue, then follow the directions for institutional log in. FAQ page for sign in
Workset: Now, we are ready to import our HT Collection into HTRC Analytics as a workset.
Click on the “Worksets” option on the header menu.
Click on Create A Workset.
Import from file:
To create a workset from a file of HathiTrust volume IDs ( in the TSV file you created in step 1) click on upload file.
Name your workset and write a description if you like.
When naming, please note that only characters A-Z, 0-9, (), -, or _ are allowed, so do not use spaces or other special characters.
Click on “Choose file” to upload the TSV file we just downloaded.
Check the “Make private workset” box if you want to create a private workset.
Finally, hit the “Create Workset” button.
If successful, you should be brought back to your worksets page, and the new workset that you just created should be listed.
Click on the name of the workset to view it.
Import from Hathitrust:
Create a workset from an existing, public HathiTrust collection
You will need the URL for a public HathiTrust collection. Private collections cannot be imported by URL.
After entering the URL, click on Fetch collection. after the collections is successfully fetched, you will see the text Data retrieved from source
Enter Name and Description, then click on Create Workset.
Algorithms: Now, we are ready to apply HTRC Algorithms to our worksets.
Click on the “Algorithm” option on the header menu.
Choose one of the Algorithms to Execute.
If you have questions about what preset choices the Algorithm is making, read the documentation.
Once you choose an Algorithm, click on execute.
It will take you to page with a description and you will have to choose a Job Name, a workset for analysis, and any additional choices relevant to the algorithm you are choosing.
Once you make your choices, you will go to a page listing current and past Jobs.
Click on the Job name to see the output.
Workset review!
How did it go?
What kind of search criteria did you use?
Did you find any challenges?
Remember you can always click on Read Documentation or Learn More to find additional contextual information.
All HTRC Tutorials#
Some content in this session based on HTRC Digging Deeper, Reaching Further used under a Creative Commons Attribution-NonCommercial 4.0 International License.