Our team is creating a Schema.org markup for legal help websites, to make them more discoverable when people search online for legal issues.
Check our current Legal Help Schema markup standards here: https://betterinternet.law.stanford.edu/markup-your-site/
People are turning to the Internet search to figure out their life problems, to determine:
- Is this situation that I’m in a legal problem? Is there anything I can do about it?
- What does the law say about it? What is it called? What do I need to know?
- Do I have any options? What exactly would I be doing to address it?
- Who can help me deal with it? How much does it cost? Am I eligible for any free or low-cost help?
Can we deliver them more correct, relevant, and actionable information on a Search Results page, or via a voice-based assistant? In a way that is prioritized, so they are given a clear, reliable path to follow?
Can we direct their natural language, non-legal searches towards the correct legal resource for them? Particularly in emergency situations around domestic violence, evictions, etc., can we connect them to the most meaningful, useful resource as quickly as possible?
We are building Schema.org markup to improve legal search
In this project, we are developing a protocol for how to apply Schema.org markup to legal help websites, and it will create and apply this markup to Florida, Hawaii, and California legal help websites as an initial case study. Schema.org is a non-profit, founded by Google, Microsoft Bing, Yandex, and Yahoo, to have a standard way to markup websites so that search engines can identify what content is on a given site page.
The goals of our initial efforts are to:
- Determine standard Legal Help Schema.org protocols, for how site administrators can consistently mark up their sites to describe the resources, referral information, and other materials on their site.
- Apply these protocols to Florida legal help websites, to encode their sites with the markup, and to ensure that it is done correctly with a technical and legal review.
- Determine if it is possible to automatically draft Schema.org markup for legal help websites, and review the results of our attempts to determine if we can propose an automated solution to be used nationwide.
- Evaluate how Schema.org markup’s application to a legal help website affects its placement on search results, and sample users’ responses to it.
Our group, in consultation with other subject matter and technical experts, will draft an initial set of Schema.org terms to use when marking up a legal help website. This draft will be an agreement of how a website will indicate what issues, jurisdiction, and organizational details are present on the site. For the most part, this draft will determine which existing Schema.org markup terms can be applied to the sites.
- For example, to indicate jurisdiction, we can use AdministrativeArea term, http://schema.org/AdministrativeArea. We can define these terms with OCD-IDs, developed as a part of the Google Civic Information APi, https://developers.google.com/civic-information/docs/v2/representatives/representativeInfoByAddress
- For legal issue area, we can use http://pending.schema.org/about, and define it by linking to the Legal Issue taxonomy’s page for relevant legal issue.
- For self-help legal guides, we can use the HowTo term, http://schema.org/HowTo . We can also tag them with Step names, Specific Directions, Estimated Time, Estimated Cost, or any other known variables about the process being described.
- For forms, we could use http://schema.org/DigitalDocument
- For legal aid organizations, courts, self-help centers, and other service providers, we can use http://schema.org/LegalService , http://schema.org/Courthouse
Potentially, the initial draft might involve the proposal of additional Schema. If discussions show that there is important content on the legal help sites that cannot be captured adequately with existing terms, our group can propose additional terms — and potentially an entire branch of Legal Schema markup — to be added to the central Schema.org standards.
We are drawing from Schema.org markup standards and examples from parallel government and civic services providers, like:
- the W3C group has a CivicServices schema plan (that integrates a proposal from the Google.org Civic team and that is profiled on this Schema.org blog page.
- the UK government’s Gov.UK site has a Schema.org plan, and
- the Dept. of Veterans Affairs had a preliminary Schema.org plan
- the US federal government has emergency (coronavirus) Schema.org strategies
We are also inspired by other projects that are about creating standardized, accessible civic data, like:
- Open Civic Data Division Identifiers (OCD-ID) to have unique identifiers for political geographical areas, with documentation here
- The Open Eligibility Project that has standards for the human services sector, from the referral group Aunt Bertha
- The EU’s Core Public Service Vocabulary data model for public, government services
- The Open Referral project to create standard referral data standards, and its Human Services Data Specification
And we have been looking at other Schema.org markup generators, at different levels of development, including:
This drafting and experimenting phase will be carried out with a mixture of subject matter and technical experts, in consultation with Schema.org community members, to ensure that our draft protocols will be durable for future uses, technically correct, and likely to be integrated into Schema.org standards.
Issues with current Legal Help search
Over the past four years, our Lab has focused on how people attempt to use the legal system to address their life problems around family (divorce, child custody, wills); around money (debt, compensation, disagreements); around housing (eviction, landlord/tenant issues); and around immigration.
Internet search is one of the primary starting points for many people, as they start to determine:
- What kind of problem their problem is — meaning, what kind of ‘system’, ‘agency’ , or ‘non-profit’ can help them. Many people do not recognize their problem immediately, as a legal one — that they have rights, access to courts, access to legal help with.
- What services they can access — free or low cost court self-help services; legal aid groups; pro bono attorneys; online legal forms, guides for DIY services; forums to discuss and strategize
- What they can do and what is ‘normal’ to do — what are recommended steps of action, how long they can expect a process to take, how much it might cost, what red flags to watch for, what would indicate a problem or a poor quality service.
In the course of our focus groups, online task observations, and our discussions with legal aid groups and courts, we have identified several failpoints in the current search engine results:
Jurisdiction incorrectly interpreted
In most help queries, the search results are not presented with jurisdiction flagged as a major criterion for the user to consider when making sense of pages’ relevance. Users also do not recognize that jurisdiction (by municipality, by state, or even by country) should condition which pages they visit and trust. We have collected reports about people in California following divorce procedure for courts in Maine or Wisconsin. A legal aid group in Australia reported this is a major problem for them: US-based resources tend to surface higher on legal help searches in Australia, and many of their clients begin their legal work using US options and procedure, so the lawyers have to ‘undo’ this work and start over in the Australian context.
Public post-disaster emergency resources out-placed by private ones
A local legal aid group in California, Legal Aid Association of California reported to us that they struggled to get public, free resources to people affected by the wildfires, mudslides, and other disasters. They are the state-appointed group responsible for maintaining all legal help information, at the statewide Law Help portal. After disasters, they set up special pages for people affected by the disaster with resources and services. Their goal is to get affected people to these free and low-cost, reliable services. They struggled with how to get people to come to their offerings, rather than private, costly services.
- During the last few months of 2017, Legal Aid Association of California, dealt with California wildfires and mudslide assistance coordination and relief efforts. As a statewide organization, Legal Aid Association of California was part of the coordination effort for statewide legal aid resources and hotline for the disaster victims. The team posted free pro bono resources and information about the emergency hotlines to the public on LawHelpCA.org, a free statewide legal aid resource, and referral website for more than 20,000 visitors a month.
- For their disaster relief outreach efforts, they posted many resource ads on their LawHelpCA Adwords Pro Nonprofit account. While they were grateful for the generous Adwords budget, they were limited by the keywords cap of $2.00 in the effort to advertise and publicize the information. During the outreach efforts, they had trouble reaching out to the victims, because they were outbid by private attorneys and solicitors who bought up the popular search terms at $4, $5 and some even at $7 per search. As a nonprofit organization and public interest group, they do not have a large marketing and advertising budget like most of these private for-profit attorneys. Because of their budget restraints, they had to go for the keyword quantity rather quality, which in terms highly limits the effectiveness of they outreach effort.
- They want to be able to find a way to work more closely with Google, Bing, Facebook, Apple and other search providers and platforms to do outreach to victims of the natural disaster to ensure that they are able to make free and reliable resources available to as many victims as possible — and to combat misinformation.
Immigration and other free government forms being charged for
Another common breakdown regards government forms. We reviewed consumer complaint websites for various for-profit legal and immigration services. In these reviews, we discovered a frequent complaint: that people searched online for how to apply for a certain visa or status, and then ended up paying a 3rd party service to fill in a form, rather than being able to fill it in for free via a government website. These 3rd party sites often have design indicators (flags, eagles, seals) that give people false confidence that they are on a government-run site, and so they invest time filling in forms and paying fees, only to discover that the fee is not a filing fee but a form-filling fee — and they must still submit this form to the government and pay a filing fee. If a person does not search explicitly for a form, but rather for a status or a process, they tend not to see the official government form place highly in the search results.
Emergency, time-sensitive queries not clearly actionable
Another concerning breakdown is around what we consider ‘emergency’ legal help searches, that indicate that there is either a very immediate deadline to take to protect a person’s rights — or that a person is in danger. In particular, these include restraining orders and domestic violence queries, and around eviction-related notices. For these types of queries, people struggle to make sense of the myriad and lengthy results. What we recommend, based on our design and focus group testing, are more directly actionable results that are given on the search results page — without having to click through many various pages to collect information and service referrals. This would mean a prioritized box on the search results page with a local hotline to call; a notice about what action to be taken; and service referral information — intake line/offices or shelters to go to/official guides form courts.
Official, free help guides not appearing high on results
Another breakdown regards how court guides place on search results. California Courts have created an extensive online self-help (DIY) resource center, http://www.courts.ca.gov/selfhelp.htm. They have the most accurate and court-sanctioned information about most non-criminal and non-immigration legal help issues in the state. The content is reliable, detailed, and linked to free in-person and online court resources. Partly because of the site’s content management system and the way the content is written, its guides are not showing up on Search Results in prominent ways.
For-profit services advertising using non-profit keywords
Legal aid groups and foundations have reported to our Lab that they have seen their names, like “Legal Services Corporation” (the Congressionally-funded civil justice funder/coordinator) and “[local legal aid group name” being used as advertising keywords by for-profit lawyers. They are extremely concerned that this use of their name then leads people to click on the ads, with the expectation that it is for free legal services or that it is verified by the government. They have reported seeing this in Maine and Massachusetts (they came across it accidentally, they have not looked more systematically).
The process that brought us here
Working Group round 1: November 2015
Our Lab scoped out the Better Legal Internet project in 2015, as Margaret Hagan was working on her paper and research on the Internet as a legal help service.
We brought together key stakeholders from legal service providers, courts, technology companies, and researchers working on legal service innovation, open data, search, and related topics in 2 summits.
Our first meeting was in November 2015, and it brought together people to scope out problems, opportunities, and specific projects to be done to improve the user experience of legal help online. Some big themes included coherent and interoperable legal portals; data standards and markup of webpages; and more work on the best design patterns and standards for online resources.
This working group began discussions of how a pilot of data standardization for one area of legal services could proceed, and then use that in generating tools and interventions for the general public’s benefit.
The session with stakeholders ended with:
- developing a shortlist for what needs to be improved for a better legal internet: including coordinated data standardization among service-providers, better user-facing tools and platforms, and other coordination of service providers & systems;
- defining what specific pilots we could implement in the Bay Area (or beyond) for a discrete legal issue, to test and gather data about these user-facing tools or back-end coordination;
- soliciting participation from a wider community to support these efforts, and promote the importance of improving the lay person’s experience of finding legal help online.
The first round of work: Semantic Schema.org markup
The primary mission that emerged from round 1 was to develop a Schema.Org extension for legal issues, procedures, and services. This extension should contain markup terms for the content that courts and legal help providers have on their websites.
As more people turn to search engines, like Google, Microsoft Bing, and Yahoo, to figure out what their problems are, and to determine how best to address them, there is an opportunity to provide meaningful, correct, and public information to them about how the legal system might be useful to solving certain of their problems — like around housing, divorce, child custody and support, cleaning a criminal record, or domestic violence.
Schema.org is a standard markup language that Search Engines all rely on, to understand what is on a give website.
- Schema.org markup can be used by groups that provide legal information online (like courts, legal aid groups, statewide legal assistance sites, and others) in order to tag up their information in machine-readable ways, about what issues and services it concerns.
- Our project will create a standard Schema.org markup protocol that can then be used by legal service providers to encode their content, so that there is a standard way to use all of these many possible terms.
- Then search engines and other applications can use this tagged-up information to find the best legal help resources for a specific user and their needs.
- It might also translate into Search Engine results pages giving more real estate or priority boxes to legal help information.
Working Group round 2
In July 2017, we hosted a 2 day workshop with a wider, national group of people working on a better legal internet to begin experimenting with what a Schema.org legal help extension might be. The goal was to build off our first round of work and priorities with many stakeholders from search engines, legal aid groups, courts, case management system and content management system providers, and data standards experts.
We spent the first day learning the details of Schema.org and its application in other fields from an expert in the markup. We then began working in smaller groups to draft possible markups to common types of offerings on legal aid and court websites, to understand how we might need to create new fields, how we can use existing fields in standard ways, and what use cases or difficulties we should be focusing on.
We finished the meeting with initial drafts of a Schema.org extension for legal help websites, which our Lab team then used to begin the tool on this website to draft information markup.
This meeting also made clear that to capture the issue-specific content on websites, to be able to link Search-users to it — there needed to be a standard way to describe the issues and resources. We discussed whether to create a new ontology of legal issues, or what we could build from. There was a consensus that the National Subject Matter Index was a robust set of terms that could be the starter ontology, but that it might need to be adjusted to be more about people’s problems rather than about service provider’s offerings.
Pilot projects in 3 states
Beginning in 2019, we have begun work on developing, deploying, and evaluating Schema.org markup for three states’ legal help websites: California, Florida, and Hawaii.
Our first round will focus mainly on describing the organizations themselves, and their main contact points and types of issue and language availability.
In later rounds of markup, we will focus on particular services, information/tools/guides, and events offered by the organization. As these change more frequently, it is worth investing in a more detailed plan to capture and present the information about these particular offerings of the organizations.