The NextMail’11 workshop will focus on current research and emerging trends in email research. I’m happy to be a part of the program committee for the workshop, which will be held as part of the IEEE / WIC / ACM International Conferences on Web Intelligence and Intelligent Agent Technology August 22, 2011 in Lyon, France.

You can read the full Call For Papers for all the details, but relevant topics include:

• Email content analysis, information extraction, summarization
• Email social networks in enterprise
• Email management strategies within organizations
• Adaptative email agents and semantic agents
• Emails archives exploration, visualization, regulations and behaviors
• Email visual interfaces and human/computer interaction with emails
• Case studies, experiments and user studies on emails usages
• Benchmark and email testing datasets
• Interoperability over email with enterprise resources and legacy systems
• Semantic email and email mining
• Unified messaging and web interactions : instant messaging, RSS feeds, annotations, tagging
• Personal information management integration in email clients, pending task management
• Interaction between email , PIM and the mobility factor
• Facing the volume growth, do we need to replace the old protocols?
• Evolution of infrastructures and uses

Papers are due by 21st March 2011, so get writing!

Launched 5 months ago Subtextual (until recently, known as bccthis) is an interesting plug-in for Microsoft Outlook that allows the mixing of public (visible to all recipients) and private (visible only to specific recipients) content within a single email message. This allows a sender to send a single message, but add private context addressed to only those people that need it.

Subtextual adds the ability to send a hidden message as part of a normal email message. This hidden content is visible only to selected message recipients – other recipients never see any indication that the message has any additional content. Happily, recipients don’t need any plug-in to view Subtextual messages.

As well as the Outlook plugin, Subtextual also have a Twitter client (which seems less compelling to me), a FireFox plug-in for using Subtextual with Gmail and a BlackBerry application.

While clearly an interesting idea, I’m not sure whether Subtextual, is significant enough to be more than just another feature for Outlook. I am, however, impressed with their family of products across desktop, mobile and web-based email. Together with their recently announced premium version of the Outlook plug-in, it feels like the company is busy experimenting, trying to discover the platforms which can deliver them traction, customers and revenue. I am very interested to see in which direction this company will pivot in the future.

The focus seems to be on lightweight interaction, which is definitely the right approach. To add a new task, for example, you just click in an empty part of the task list and start typing. This seems pretty similar to the style of task interaction pioneered by Remember the Milk, and I’d be interested to know how it compares with RTM’s GMail services, particularly their recently announced RTM GMail gadget that can be added via GMail Labs. Are there any users out there who have experimented with RTM’s tools and can offer insight on the comparative strengths and weaknesses of the new Labs task addition?

There doesn’t seem to be much in the way of tight interaction between email and tasks (yet), but I’m sure this will be in the pipeline for future enhancements.

On the topic of tasks in email, if you’re interested in learning more about how people phrase tasks in email messages, have a look at my recent paper, Requests and Commitments in Email are Complex: Eight Reasons to be Cautious, which I presented at the Australasian Language Technology conference in Hobart earlier this week.

Parts of my on-camera conversation with Kate have made it into the recently released HCSNet Promotional video, which is now available on YouTube. It’s always a bit weird seeing yourself on camera, particularly when sound bytes are taken from a much longer conversation! Given the totally unscripted nature of what was recorded though, I think it’s worked out quite well.

Of course, this is also a good opportunity to actually plug the annual HCSNet Summerfest, which will be held at UNSW in Sydney in December. If you’re interested in speech, language, sonics, psychology or any number of topics in between, check out what’s on offer – it’s well worth a few days of your time to meet some inspiring people.

Brij thinks that the application of sentiment analysis to email could help address this mental wandering for knowledge-based employees:

I think it’s high time for companies to invest in sentiment classification and routing toxic emails to platform where immediate impact on employee productivity is less. Can carefully controlled social platform enable this process?

Having just yesterday attended a research presentation by Mary Gardiner on sentiment classification, it’s interesting to consider the possibilities and practicalities of applying the sentiment classification techniques to email.

One unsupervised technique, pioneered by Turney and Littman, is to use pointwise mutual information (PMI) and word co-occurrence counts from a search engine to help determine the valence of each word in a text. Turney and Littman used the NEAR operator in Altavista to determine the co-occurrence of each word in their text to be classified (in our case, this would be each word from an incoming email message) with each word from a set of words with known positive or negative valence. The counts for co-occurrence with the known-positive words contribute to the positive sentiment of our unclassified word, while counts for co-occurrence with negative words contribute to the negative sentiment. These co-occurrence counts are then normalised and combined to determine the overall valence of each word from our unclassified text. The technique, though simple, worked surprisingly well (80% classification accuracy at the word level), much better than many more complex techniques.

Ignoring the sad reality that the NEAR operator is no longer available to use in Altavista queries (and that no other search engines offer an operator of similar functionality in their public query interface), it’s interesting to think about whether such a technique could be usefully applied to email. I don’t know if people have addressed how to move from word-level classification up to message-level sentiment classification, but it doesn’t seem to be an insurmountable problem.

More of an issue for email is whether people would be happy for the entire text of their email messages to be sent in clear text to a single search provider. Depending on the volume and nature of data on a user’s own machine, perhaps we could use the desktop search interface to approximate Turney and Littman’s technique, without passing sensitive email data out onto the network? Of course, there’s a big difference in the scale of corpus being used to generate the co-occurrence counts in this case – Altavista at the time of the experiment, claimed to be indexing around 100 billion words. My desktop search index claims to contain about 1.5 million items (email messages, documents, visited web pages etc.) . While that’s not going to get us to 100 billion words, it might be enough to get some credible results?

For people interested in Xobni’s work on displaying person profiles directly in people’s mail client, they’ve recently added further functionality to their Outlook plugin.

Fuser and Orgoo both focus on the integrated/universal messaging client, bringing IM, social networks and other communication mediums into a single client along-side email. Xoopit is still in stealth-mode, so they haven’t revealed much publicly about the details of their work, but their focus appears to be on extracting and compiling collections of attached documents, images etc. from email archives. More interesting to me is Xobni, who I’ve been following with some interest since Vitor Carvalho brought the company to my attention a few weeks ago.

Chris Morrison notes that while Xobni already pulls out some information like phone numbers from email, there’s much more information waiting for someone to find an innovative way to highlight. Of course, highlighting is only one option for making such structure available and useful for end users. Matt Brezina, co-founder of Xobni, also comments about the latent, untapped structure in email:

“There’s a structure that just hasn’t been broken apart and exposed”
Matt Brezina – Co-Founder Xobni

I think Matt is right on target with this assessment. It’ll be interesting to see which avenues of structure they pursue. I have my own ideas on important latent structure in email, some of which you can hopefully read about in an upcoming conference paper. More details coming if and when the paper actually gets accepted.

Anyway, if you’re at all interested in human factors, artificial intelligence or language technology, take a moment to tune in – we have some excellent talks coming up in the near future. As you can see from our collection of past seminars, topics range widely including research and applications in usability, human-computer interaction, user modelling/personalisation, novel interfaces, natural language processing, linguistics, information retrieval, speech processing, system evaluation, computer supported cooperative work, cognitive science and more.

On September 19th, I will be presenting a seminar to the NSW branch of CHISIG - the Computer-Human Interaction Special Interest Group of Australia – about our research in CSIRO that focuses on controlling the flow of information to deliver the right content to the right people at the right time in the right form.

Our research approaches the problem by using knowledge about users and their interaction to tailor the information that is gathered and to present it appropriately. The context information that is captured and reasoned about can include user preferences and characteristics, as well as details of a user’s current task, their previous history of interaction and their environment. This context can determine which information should be retrieved, and how that content should be aggregated, organised, and presented, in order to best support the user.

My presentation will cover work that builds on concepts and techniques from a variety of different fields, including: natural language generation, information extraction, information retrieval, discourse analysis, user modelling, task analysis and HCI, so if any of those topics spark interest (and you happen to be in Sydney) you might consider coming along to PTG Global on Tuesday 19th.