Topicmarks is a web service that creates smart, interactive synopses from electronic documents in minutes.

We use the most advanced natural-language processing technology available to understand the true meaning of your text, summarize it for you and retain all its facts forever.

Upload multiple texts and Topicmarks becomes your knowledge base: finding links between texts, grouping texts around topics, and containing all the facts from all documents you ever read.

Understand the essence and retain every single fact - without reading anything yourself.

Word processors revolutionized writing. The web revolutionized publishing. But where are the tools to revolutionize reading?

To conquer information overflow, we need to let computers help us with reading too.

Topicmarks helps you create smart, interactive synopses from electronic documents, that contain all original information yet give an immediate overview of the essential contents.

This helps you get the gist in minutes without having to read the full text yourself.

What's more: Topicmarks extracts all facts from texts you read and stores them in your own personal knowledge base, so that you will always remember any single fact you have ever "read".

With Topicmarks, you can catch up with all your reading, without ever having to read anything yourself.

Topicmarks works by cutting up the full text into sentences, attaching meaning like synonyms and antonyms to each word in the sentence, and then by identifying triplets (typically between subject, verb and object) that are the basic facts expressed in the text.

Topicmarks is useful for anybody who needs to read long or many electronic documents. And who doesn't?

Topicmarks appeals especially to people who have to read and study documents as part of their main occupation. We have many:

• Academic researchers
• Authors
• Bloggers
• Business analysts
• Consultants
• Executives
• Graduate students
• High school students
• Intelligence functions
• Journalists
• Lawyers and Paralegal staff
• Managers
• Personal assistants
• PhD students
• Teachers and Professors
• Technology researchers
• Undergraduate students
• Watchdogs
• Writers

Summarizing text and storing its facts forever has many applications, depending on your field of work.

For example, we see many customers using Topicmarks intensely for:

• Analyzing your own writings
• Being up-to-date with the latest financial research
• Building a knowledge base for a graduate thesis
• Building a knowledge base for a master's thesis
• Checking current facts against past press releases
• Checking doctoral theses
• Discovering emerging patterns
• Evaluating student papers quickly
• Finding back quotes they remember having read somewhere
• Finding inconsistencies in long reports
• Getting the gist of subordinates' presentations
• Preparing a school project
• Reading up quickly on industry analyses
• Researching a first student paper
• Sifting through annual reports
• Sifting through legal cases
• Staying abreast of white papers
• Storing relevant legal precedents
• Writing fiction abstracts y analyses
• Researching a first student paper
• Sifting through annual reports
• Sifting through legal cases
• Staying abreast of white papers
• Storing relevant legal precedents
• Writing fiction abstracts

Currently, Topicmarks supports the following file types:

• Adobe Acrobat PDF (.pdf)
• Microsoft Word (.doc, .docx)
• OpenOffice ODF text (.odt)
• Hypertext Markup Language (.htm, .html)
• Plain text (.txt)

We will be adding more file formats to Topicmarks in the future.

For now, Topicmarks is designed to work only with English texts. You will get an error if you try uploading a text that is in another language.

Machine reading and natural language processing are very language-specific. Even related languages are so different in their subtle meanings that it is near impossible to write software that can deal with multiple languages at once. This is why we designed Topicmarks to work properly only with English texts. And while this team of language geeks would never exclude adding other languages in the future, we will most likely only be able to finance other languages once the English version has taken off massively. In the alpha and early beta stages, Topicmarks did not yet filter for English and some users reported moderate to good results in other languages. We found the quality to be mostly wanting, though, so we decided to allow English texts only until we can provide a better quality in other languages.

We thank you for your understanding - Merci de votre compréhension - Agradecemos su comprensíon - Danke für Ihr Verständnis - Met dank voor uw begrip

If you have properly summarized/condensed a text, then Topicmarks will store your text forever. We do not charge for this storage for now, although we do not exclude charging a nominal fee for heavy users in the future.

If you have uploaded a text and then clicked "discard", Topicmarks will not store your text. You will have to upload it again if you want it summarized or condensed.

By default, texts uploaded to Topicmarks are visible to friends. This means that they are available to search engines and that you can share the URL of the text knode with friends over Facebook, Twitter, email or other social media.

If you want to ensure complete confidentiality, go to the "Properties" tab of your text knode and switch it to "private". This setting means that only you can review the text knode and that it is not available to search engines anymore. Of course, it t also means that any friends that you might have shared the text knode with, cannot access it anymore.

Sharing a text knode with your Facebook friends is simple.

1. While viewing the text knode, click on the Facebook icon in the bottom left.
2. You may be required to log in to your Facebook account
3. On the "Post to Profile" page that appears, add a personal comment for your friends in the text box
4. If you prefer to send the text knode to a friend privately, click on "Send as a private message instead" in the bottom left.
5. Then click on "Share"
6. Sharing text knodes requires them to be visible to friends. If yours was private, we will switch it to "visible to friends" and warn you about it.

Sharing a text knode with through email is simple.

1. While viewing the text knode, click on the email icon in the bottom left.
2. In the "Email a Friend" dialog that appears, add your friend's email address, your own email address and a personal comment for your friends in the "Note" box
3. Then click on "Send"
4. Emailing text knodes requires them to be visible to friends. If yours was private, we will switch it to "visible to friends" and warn you about it.

Most people compare Topicmarks to other knowledge base services like Evernote and Zotero, but our focus is really different.

Evernote focuses on storing and making accessible as many notes and documents as possible. It is excellent in reading multiple document formats and in making scanned or photographed documents electronically readable (OCR), which is why we often recommend it to people wanting to upload a scanned document to Topicmarks. But unlike Topicmarks, Evernote does not interpret the content, extract facts and synonyms, or summarizes texts for you.

Zotero is primarily focused on managing academic references, although it is possible to upload original copies of the texts into your library. Zotero is also mostly a Firefox plugin, whereas Topicmarks is a pure web-based "software as a service" (at least for now). But the most important difference is again in the content interpretation. Unlike Topicmarks, Zotero does not interpret the content, extract facts and synonyms, or summarizes texts for you.

You can get a text into Topicmarks in four ways:

• Uploading a file from your hard drive
• Copy-pasting text directly from an open document into Topicmarks
• Letting Topicmarks download a document from a URL you give it
• Sending a file to Topicmarks by email (check the next question for details)

1. Go to your dashboard by clicking on "My text knodes" (top right, second line)
2. Check if the Upload dialog is already open. If not, click the link "+ upload text" next to the "Explore:" search field
3. Select the relevant tab: upload a file, paste a text or paste a URL
4. For "upload a file", click the browse button and select the file to upload from your hard drive. Then click the big blue "Upload" button.
5. For "paste a text", go to your word processor or browser and copy the text you want to analyze, then come back to Topicmarks and paste the text in the text area field. Then click the big blue "Upload" button.
6. For "paste a URL", go to your browser and copy the URL of the page you want to analyze from the "Location" field. Make sure it is a full URL starting with "http://", "ftp://" or something similar, and paste this into the Topicmarks URL field. Then click the big blue "Upload" button.
7. Follow the instructions on the subsequent screen.

Yes, you can! So many of the files we have to read come in by email. Now you can just forward these incoming messages to your own special email addresses and Topicmarks sends you back summary. Fully integrated in your work flow, you don't even have to visit the Topicmarks Web site again!

To send files to Topicmarks by email, you just need the special email address that is linked to your account. Log in to Topicmarks, click on "profile" and look for "your email-in address". It looks like this: 1B56A051-E8F7@topicmarks.com . This address is linked only to your Topicmarks account and we have made it difficult to guess in order to keep spammers out.

Copy the address and add it to your address book, giving it a name like "My Topicmarks" or "Kim Topicmarks" (if your name happens to be Kim). Now you can just forward any incoming email to "My Topicmarks" resp. "Kim Topicmarks" and Topicmarks will send you back a summary in a few minutes.

Please note that Topicmarks summarizes the first text attachment it finds in your email, or the body of the text if it doesn't find any attachment. In the future, we will expand this functionality so that Topicmarks can summarize multiple attachments per message plus the body text.

Text knodes are designed to give you a thorough understanding of the underlying text, no matter how long, in just four minutes:

1. Go to your document dashboard and click on the text knode you want to review
2. Click on the "Summary" tab and read the summary. Expand and reduce the summary to find the key phrases (1 min)
3. Click on the "Keywords" tab and review the tag cloud. Click on any keyword of particular interest (1 min)
4. Click on the "Facts" tab and review the extracted facts. Sort facts and click on any that is new to you (1 min)

5. Click on the "Index" tab and review any keywords you may have missed. Expand and reduce to find key concepts (1 min)

6. Synopsis fascinating? Then by all means, read the original!

7. Not fascinated? Save your time and do something else. The text knode and original remain safely on your document dashboard, and all its underlying facts are stored in your knowledge base forever.

Using the power of the semantic web, Topicmarks tries to extract every fact your text contains. As a starting point, it does this in the form simple subject-verb-object triples like "My son" "likes" "cookies", with more advanced fact interpretation coming in the future.

The facts view of a text knode gives you access to all the facts extracted from your text. Depending on the size of the text, these may be just a few or hundreds and hundreds of individual facts, that will always be stored in your personal Topicmarks knowledge base.

1. Click on the "Facts" tab of any text knode
2. Click on "Weight" to sort the extracted facts by importance (reflecting the overall content of the document)
3. Scroll through the facts displayed and browse further pages (bottom right) to see all the facts
4. Search for a specific fact by typing one or two words in the Search box (top left of the text knode, right under the Facts tab)
5. Share the text knode with your Facebook friends or Twitter followers by clicking the appropriate buttons
6. The message to your friends will contain a link that allows them to review the fact list of your text knode for themselves.

Using the power of the semantic web, Topicmarks tries to compress your text into the key words and sentences that your document contains. You can dynamically adjust how many sentences the summary should contain, or which specific items you want the summary to focus on.

1. Click on the "Summary" tab of any text knode
2. Review the summary. You can highlight the weight of any sentence in the overall text by hovering over it with your mouse, and of any keyword by hovering over the underlined keywords.
3. Use the slider (top right) to dynamically extend or shorten the summary (typically from 3 to 50 sentences)
4. Click on a specific keyword to summarize the text subjectively around that keyword. You can also enter a specific keyword in the "Search:" box (top left) to accomplish the same effect. Use the "strict" checkbox to get only sentences that contain your keyword
5. Share the text knode with your Facebook friends or Twitter followers by clicking the appropriate buttons
6. The message to your friends will contain a link that allows them to play with the summary of your text knode for themselves.

Using the power of the semantic web, Topicmarks tries to index all important keywords that your text contains. You can dynamically adjust how many entries the index should contain.

1. Click on the "Index" tab of any text knode
2. Review the index. You can jump directly to any character by clicking on the alphabetic list at the top.
3. Use the Coverage slider (top right) to dynamically extend or shorten the list of index entries (typically from 10% to 100% coverage)
4. The numbers behind every keyword refer to the number of the fact in which the keyword occurs. You can click on any fact number to review that fact directly in the facts view.
5. Share the text knode with your Facebook friends or Twitter followers by clicking the appropriate buttons
6. The message to your friends will contain a link that allows them to play with the index of your text knode for themselves.

Using the power of the semantic web, Topicmarks tries to extract all important keywords that your text contains, and display them in a keywords swarm (or tag cloud). You can dynamically adjust how many entries the swarm should contain.

1. Click on the "Keywords" tab of any text knode
2. Review the keywords swarm. The keywords in a larger font are the more important ones. Use the search field in the top left to limit the swarm to certain keywords only.
3. Use the slider (top right) to dynamically extend or shorten the number of keywords that the swarm contains (typically from 20 to 150)
4. You can click on any keyword to review the relevant facts directly in the facts view. This is the same effect as searching the fact view for that keyword.
5. Share the text knode with your Facebook friends or Twitter followers by clicking the appropriate buttons
6. The message to your friends will contain a link that allows them to play with the keyword swarm of your text knode for themselves.

A "text knode" is Topicmarks' name for the smart synopsis you produce when summarizing a text using our technology.

We have created the word "knode" as a contraction of "knowledge" and "node".

So we should spell it "knowledge node" if we wanted to stick to standard English, but we liked "knode" better. We hope you can forgive us.