Frequently asked questions
- Who has access to Link?
- Only faculty, staff, and GSAS students with active HarvardKey accounts have access to Link.
- How can I control what Link displays about me?
- Users can remove their websites, which will drastically reduce their footprint in Link. Once a website is removed, the user will only be recommended to others based on their taught courses' descriptions (if any).
- Faculty can opt out of the Syllabus Explorer project, which would also mean their syllabi would not be available on Link.
- How can I improve my recommendations?
- Why is my personal information incorrect and/or missing?
- If any of your information (name, email, website) is incorrect or missing, please update it now. The site compiles your information through a combination of data sources that are either public or internal to Harvard, not all of which may be complete or accurate. Keeping your information updated will improve Link's services for you and other users.
- Why is the site different from what my colleague sees and/or what I see in the screencast?
- Link will populate site content based on the data we have about you. For example, if your school has not yet agreed to join the Syllabus Explorer project, then you will not receive recommendations for courses with similar syllabi. If you believe we should have data on you that is not displaying, please send us a message.
- How frequently is data on the site updated?
- The site is updated daily through direct connections to the data sources.
- I just updated my website or keywords. When will I start getting recommendations?
- We will analyze your website's text and incorporate this information into your recommendations on the following day. If we don't have a website URL for you and you've never entered any keywords, the first time you add keywords we will use them to populate some temporary recommendations until the following day, when we will use all information available (including your website's text) to recommend content to you.
- How are the recommendations being made?
- Recommendations on related people, news, and events are powered by a textual analysis of the data sources associated with each user (e.g. an OpenScholar website) in comparison to other people, news articles, and event listings. If you have an event recommended to you (that is coming from a publicly available event feed), and are unsure if you are allowed to attend, please contact the event organizer.
- How does Harvard Link get its data?
Data type Source Personal information (e.g. name, email, positions) Harvard Data Warehouse Public websites for faculty, staff, students, and organizations OpenScholar API and web scrapers for non-OpenScholar sites Syllabi Canvas course sites Course names/descriptions my.harvard Student Information System and Operational Data Store Quick links Curated by VPAL and assigned based on affiliation News and events Various Harvard RSS feeds
The above data is primarily collected through direct access to databases. For websites that cannot be accessed directly, Link collects them through a technique called "web scraping", whereby the contents of a website are programmatically collected on a periodic basis. All of our data is cleaned and formatted so that Link can analyze and display the data in a safe and consistent manner.
We encourage faculty and other researchers who have not yet set up an OpenScholar website to do so. It is a free, easy-to-use service made available the Harvard community. You can learn more at OpenScholar at Harvard.
- How personal/private is the data being used in Link?
Most of the data used is publicly available:
- Names/emails are available in the public Harvard directory
- Websites are all publicly accessible
- Syllabi are restricted to schools that have consented to participation in the Syllabus Explorer project. To participate in that project, the school needs to authorize sharing syllabi. Individual faculty members at consenting school retain the ability to opt out of providing their syllabi. If a faculty member opts out of the Syllabus Explorer, they will also automatically opt out of sharing their syllabi on Link.
- Most of the data used is publicly available:
- What is Link doing to make sure that the data about me will be used responsibly?
- Regarding the use of syllabi, the Syllabus Explorer policies also apply here.
- Data will never be shared with or sold to parties outside of Harvard.
- User data (i.e. how users are using the site) will only be used to improve the services provided on Link. This data will not be used for any other purpose.
- Data is processed by VPAL-Research's secure data pipeline.
- As with the Syllabus Explorer project, use of data on Link will not be used for tenure or promotion decisions. With respect to faculty, the same Acceptable Use Policy for Syllabus Explorer applies to Link.
- Link collects logs that can be used to identify certain improper behavior, such as the programmatic downloading or scraping of information, which is prohibited.
We support several advanced text search features for searching news, events, people, and organizations.
- genome editing
- either of the words genome and editing, including variations of each such as genomes and edit
- +genome +editing
- both of the words genome and editing, including variations
- "genome editing"
- the exact phrase genome editing
- genome -editing
- the word genome but not editing
- genome -"genome editing"
- the word genome but not the phrase genome editing
- any word beginning with biotech
To match both hematology and haematology, search hematology~ (with the trailing tilde).
Parts of your query can be weighted higher than others. For instance, search "global health" nutrition^2 to return records about global health that often feature the word nutrition.
All of these described search features can be used within groups. For instance, +(politics OR policy) +health will match records containing both health and either politics or policy.