Streamspace

idea:A research Dashboard

smitashree — Tue, 16 Mar 2010 22:49:09 +0000

if reader is studying a paper, the system should extract the strctures and suggest the relatedarticles, authors. it should extract the citation.

it should extract the self citation and record as part of the literature review with user comments

idea:semtweet

smitashree — Tue, 16 Mar 2010 22:46:24 +0000

identifying and determining linked data entity from user tweets. Acess the hash tags, topics and links from the tweets

Keynote by tom Mitchell at ISWC 2009

smitashree — Wed, 28 Oct 2009 13:58:34 +0000

keynote by Tom Mitchell about populating semantic web using machine Learning

Points made:

manually enter by users
convert from Database
computer reading of unstructured data:

         use an initial ontology
         leverage redudancy available on web
        try to target the ontology to populate
        semi-supervised learning : couples semi-supervised learning (bootstrap learning) for category and relation    extraction.
        use seed examples from Freebase or dbpedia.

here comes o my mind can we populate existing LSCOM ontology from youtube data ? this can be good candidate for the next paper.

Lightening Talk of SDOW

smitashree — Sun, 25 Oct 2009 21:24:04 +0000

twitris: trend tracking of Twitters topics with time and space dimensions. Won semantic web challenge didn’t work very well during demo (meena Nagarajan)

personal Notes to social interlinked data: SemNote which annotates the entities automatically with URIs from existing desktop data and publishes on web as part of the linked data. (tudor)

callimachus: sw applications made easy:Object-RDf mapping (david Wood)

Freemix presentation

smitashree — Sun, 25 Oct 2009 20:37:21 +0000

Freemix: Social Networking Meets Data presented by David Wood on preserving and presenting digital data (of cultural significance) in different way. Freemix used SIMILE from MIT to visualize the metadata. presenting different formatted data upto the user needs and requirement has to be addressed urgently. We simply cannot speak about the semantic data RDF, RDFa and others in its raw form.

Discussion for interesting topics for user generated content and related research

smitashree — Sun, 25 Oct 2009 16:52:36 +0000

John Breslin:real time search in microblogging data; what sw community can do? and explore the existting and the emerging semnatics ? (john)

Amit Seth: content is more accessible in web search as compared to social web. The network effect is more complex in social web. so we need to describe the links with lot more attributes. the connection dynamism on twitter is very different from normal we is very different so the notion of search will be more personalized information and specific to users. So the challenges are more complex.

Meena nagarajan: to exploit social context and recommender system is an good option.

Challenge: Evaluation of social data:

Need more data for evaluation

why social data is only user profiles rather it also focus on group dynamics, social issues (chris).

Challenge : Opportunities and Challenges associated with interdisciplinary research : (suggested by Amit Seth)

Challenge :Rebuilding social web application using semantic web technology

it is still very hard to formulate the social interactions in RDF easily. we need more user freindly data model from semantic web

IYOUIT:Mobile context awareness service

smitashree — Sun, 25 Oct 2009 16:04:25 +0000

context based mobile user profiling using data mining techniques. IYOUIT sense all the info (tag cloud,location,weather,presence, physical env)around the user and aggregates to form the places of interest and mapped to the ontology concepts. Data mining for context corelation and used association rule mining to deduce context to build dynamic user profile.

How dot they get physical env and the presence data ?

ok the presence data is based on the status but how about likes and dislikes ? (Asked by Amit Seth) and the answer was that they access the web services available where the user gives explicit information about themselves. How do they identify as context (asked by Srikant ,Microsoft) ans :- any data about the user is considered as context.

NoTube BeanCounter talk by Chris van Aart

smitashree — Sun, 25 Oct 2009 14:06:07 +0000

Chris van Aart talked about NoTube project and the extraction of user data such as attention data,user profile,user context and domain data for rcommending television programme from social web. talked about semantic enriching of TV data from extenal sources such as IMDB.

keynote talk by Meena Nagarajan

smitashree — Sun, 25 Oct 2009 13:44:08 +0000

Introduction by John some interesting papers on “social data onweb” worksop. Discussion panel to suggest new topics for social data.http://tinyurl.com/sdow2009

Meena’s KN talk on “user generate content on social media challenges and opporunities ” talked about making up contextual data whene there is less metadata from various existing data such as social spatial and temporal dimensions, sense identification of cultural NER and the needs to move from closed world assumption from open world assumption to reduce the entity extraction complexities.

user intention mapping by means of three elements,information seeking,information sharing and

Zemanta Plugin

smitashree — Mon, 12 Oct 2009 14:47:43 +0000

Downloaded zemanta IE plugin after watching video in videolectures.net. The tool suggests images and tags to the blog post. I tried in my last blog post where it took the keywords from the post suggested as tag such as “DERI” expanded with “Digital enterprise research institute”,”sindice”,”semantic web” but the tag “ iswc09″ was not recognised ,is a bit strange though it was part of the title and tagged with. It looks like it takes the clean keywords and the strongly corelated words. regarding image It suggested “semantic web stack” and “linked data” image.

If it works it is an interesting idea to test and adapt as well. Can something be thought of for user videos ? Once the user upload videos is any tools there to suggest the tags, blog posts and news items related to the videos ?