![\"\"](\"https:\/\/world.edu\/wp-content\/uploads\/2017\/06\/AI.jpg\")
<\/p>\n<\/p>\n
We humans are swamped with text. It\u2019s not just news and other timely information: Regular people are drowning in legal documents. The problem is so bad we mostly ignore it. Every time a person uses a store\u2019s loyalty rewards card<\/a> or connects to an online service, his or her activities are governed by the equivalent of hundreds of pages of legalese<\/a>. Most people pay no attention<\/a> to these massive documents, often labeled \u201cterms of service,\u201d \u201cuser agreement\u201d or \u201cprivacy policy.\u201d<\/p>\n These are just part of a much wider societal problem of information overload. There is so much data stored \u2013 exabytes<\/a> of it, as much stored as has ever been spoken by people in all of human history \u2013 that it\u2019s humanly impossible to read and interpret<\/a> everything. Often, we narrow down our pool of information by choosing particular topics or issues to pay attention to. But it\u2019s important to actually know the meaning and contents of the legal documents that govern how our data is stored and who can see it.<\/p>\n As computer science researchers, we are working on ways artificial intelligence algorithms could digest these massive texts and extract their meaning, presenting it in terms regular people can understand<\/a>.<\/p>\n Can computers understand text?<\/strong><\/p>\n Computers store data as 0\u2019s and 1\u2019s \u2013 data that cannot be directly understood by humans. They interpret these data as instructions for displaying text, sound, images or videos that are meaningful to people. But can computers actually understand the language, not only presenting the words but also their meaning?<\/p>\n One way to find out is to ask computers to summarize their knowledge<\/a> in ways that people can understand and find useful. It would be best if AI systems could process text quickly enough to help people make decisions as they are needed \u2013 for example, when you\u2019re signing up for a new online service and are asked to agree with the site\u2019s privacy policy.<\/p>\n What if a computerized assistant could digest all that legal jargon in a few seconds and highlight key points? Perhaps a user could even tell the automated assistant to pay particular attention to certain issues<\/a>, like when an email address is shared, or whether search engines can index personal posts. Companies could use this capability, too, to analyze contracts or other lengthy documents.<\/p>\n To do this sort of work, we need to combine a range of AI technologies, including machine learning<\/a> algorithms that take in large amounts of data and independently identify connections among them; knowledge representation<\/a> techniques to express and interpret facts and rules about the world; speech recognition<\/a> systems to convert spoken language to text; and human language comprehension<\/a> programs that process the text and its context to determine what the user is telling the system to do.<\/p>\n Examining privacy policies<\/strong><\/p>\n A modern internet-enabled life today more or less requires trusting for-profit companies with private information (like physical and email addresses, credit card numbers and bank account details) and personal data (photos and videos, email messages and location information).<\/p>\n These companies\u2019 cloud-based systems typically keep multiple copies<\/a> of users\u2019 data as part of backup plans to prevent service outages. That means there are more potential targets \u2013 each data center must be securely protected both physically and electronically. Of course, internet companies recognize customers\u2019 concerns<\/a> and employ security teams to protect users\u2019 data<\/a>. But the specific and detailed legal obligations they undertake to do that are found in their impenetrable privacy policies. No regular human \u2013 and perhaps even no single attorney \u2013 can truly understand them<\/a>.<\/p>\n In our study, we ask computers to summarize the terms and conditions regular users say they agree to when they click \u201cAccept\u201d or \u201cAgree\u201d buttons for online services. We downloaded the publicly available privacy policies of various internet companies, including Amazon AWS<\/a>, Facebook<\/a>, Google<\/a>, HP<\/a>, Oracle<\/a>, PayPal<\/a>, Salesforce<\/a>, Snapchat<\/a>, Twitter<\/a> and WhatsApp<\/a>. Our software examines the text and uses information extraction techniques<\/a> to identify key information specifying the legal rights, obligations and prohibitions identified in the document. It also uses linguistic analysis<\/a> to identify whether each rule applies to the service provider, the user or a third-party entity, such as advertisers and marketing companies. Then it presents that information in clear, direct, human-readable statements.<\/p>\n For example, our system identified one aspect of Amazon\u2019s privacy policy as telling a user, \u201cYou can choose not to provide certain information, but then you might not be able to take advantage of many of our features.\u201d Another aspect of that policy was described as \u201cWe may also collect technical information to help us identify your device for fraud prevention and diagnostic purposes.\u201d<\/p>\n
\n
\nSummarizing meaning<\/strong><\/p>\n