Goto

Collaborating Authors

Results


Towards Personalized and Human-in-the-Loop Document Summarization

arXiv.org Artificial Intelligence

The ubiquitous availability of computing devices and the widespread use of the internet have generated a large amount of data continuously. Therefore, the amount of available information on any given topic is far beyond humans' processing capacity to properly process, causing what is known as information overload. To efficiently cope with large amounts of information and generate content with significant value to users, we require identifying, merging and summarising information. Data summaries can help gather related information and collect it into a shorter format that enables answering complicated questions, gaining new insight and discovering conceptual boundaries. This thesis focuses on three main challenges to alleviate information overload using novel summarisation techniques. It further intends to facilitate the analysis of documents to support personalised information extraction. This thesis separates the research issues into four areas, covering (i) feature engineering in document summarisation, (ii) traditional static and inflexible summaries, (iii) traditional generic summarisation approaches, and (iv) the need for reference summaries. We propose novel approaches to tackle these challenges, by: i)enabling automatic intelligent feature engineering, ii) enabling flexible and interactive summarisation, iii) utilising intelligent and personalised summarisation approaches. The experimental results prove the efficiency of the proposed approaches compared to other state-of-the-art models. We further propose solutions to the information overload problem in different domains through summarisation, covering network traffic data, health data and business process data.


#cloudcomputing_2021-06-09_07-26-57.xlsx

#artificialintelligence

The graph represents a network of 1,360 Twitter users whose tweets in the requested range contained "#cloudcomputing", or who were replied to or mentioned in those tweets. The network was obtained from the NodeXL Graph Server on Wednesday, 09 June 2021 at 14:35 UTC. The requested start date was Monday, 07 June 2021 at 00:01 UTC and the maximum number of days (going backward) was 14. The maximum number of tweets collected was 7,500. The tweets in the network were tweeted over the 4-day, 5-hour, 24-minute period from Wednesday, 02 June 2021 at 18:36 UTC to Monday, 07 June 2021 at 00:00 UTC.


[D] [R] BERT Base. Choosing optimal cloud infrastructure and environment setup (long read)

#artificialintelligence

Then I began increasing batch size and the number of data loader workers to maximize GPU utilization. To illustrate the approach below are 3 graphs of GPU utilization for P100.


At Think conference, IBM puts AI and hybrid cloud to work - SiliconANGLE

#artificialintelligence

IBM Corp. is pushing the envelope on hybrid cloud and artificial intelligence with a number of key announcements early Tuesday ahead of its Think 2021 event, chiefly aimed at accelerating its customer's digital transformation strategies. One of the main highlights of today's announcements is a new Auto SQL capability within IBM's Cloud Pak for Data offering that automates data access and management without needing to move it first. The company also unveiled a new, AI-based tool for modernizing applications and workloads to run in hybrid cloud environments, plus new AI capabilities in Watson and advancements that should help to scale up quantum computing to more use cases. Available Tuesday, the new AutoSQL capability for IBM Cloud Pak for Data is a big deal because it enables companies to automate access, integration and management of their data no matter where it resides, the company said. IBM said it's addressing one of the most critical pain points customers face as they attempt to reduce the complexity of curating data for AI.


AWS for Machine Learning -- Part 1

#artificialintelligence

Before the concept of cloud computing came into the picture back then even if a website needs to be hosted companies had to buy huge servers and maintain them. It was a huge cost and inefficient workforce diversion for the companies which wanted to focus on the actual task at hand rather than the maintaining of these servers. Some other companies saw this as an opportunity which went ahead and bought these huge servers and had a huge collection of servers and rented them out to other companies. It is a win-win for everyone since it is cheaper and easier for the companies that wanted to focus on their application/product rather than maintaining these servers. We all use electricity, how do we pay for this we pay according to the number of units used.


#cloudcomputing_2020-12-14_04-51-01.xlsx

#artificialintelligence

The graph represents a network of 1,522 Twitter users whose tweets in the requested range contained "#cloudcomputing", or who were replied to or mentioned in those tweets. The network was obtained from the NodeXL Graph Server on Monday, 14 December 2020 at 13:04 UTC. The requested start date was Monday, 14 December 2020 at 01:01 UTC and the maximum number of days (going backward) was 14. The maximum number of tweets collected was 7,500. The tweets in the network were tweeted over the 2-day, 2-hour, 17-minute period from Friday, 11 December 2020 at 22:43 UTC to Monday, 14 December 2020 at 01:00 UTC.


Widespread Amazon cloud service outage disables Roombas, Ring doorbells, even Christmas lights

Daily Mail - Science & tech

A large-scale outage on Amazon's cloud service this week wreaked widespread havoc on websites and software services. In addition to disabling Flickr, Adobe and the Washington Post's website, the outage of Amazon Web Services (AWS) on Wednesday caused Roombas, Rokus, Ring doorbells and other smart household appliances to stop functioning. The issue, which impacted the US East-1 region, sent Twitter users into a tizzy. 'My f---ing doorbell doesn't work because AWS us-east-1 is having issues,' tweeted one disgruntled Ring user. 'I... can't vacuum... because us-east-1 is down,' complained Geoff Belknap, Chief Information Security Officer for LinkedIn.


Towards Edge Computing

#artificialintelligence

What is edge computing vs. edge data centers? Edge computing is the activity of processing and storing data close to where the data is generated and used. Edge data centers are the physical structures where edge computing takes place and are usually located within a few miles from where the data is generated. In the examples above, the edge data center could be on-site at the hospital or at a nearby cell phone tower. Edge data centers are a part of the broader connectivity ecosystem and operate in collaboration with the central data centers.


Google Parent's Stock Soars on Gangbuster Earnings

WSJ.com: WSJD - Technology

Google blasted through the coronavirus pandemic with gangbuster earnings, just a week after U.S. prosecutors sued the company for operating a purported illegal monopoly in its flagship search business. Alphabet Inc. reported a third-quarter profit of $11.2 billion, well outstripping analyst estimates. As importantly, digital advertising revenue of $37.1 billion was up compared with last year, marking a turnaround from a quarter earlier, when the company recorded the first drop in the category in company history. Cogs across the Alphabet empire were clicking. Helped by stay-at-home trends, YouTube pulled in more than $5 billion in advertising for the first time, gaining 32% over the same period a year earlier.


#cloudcomputing_2020-10-26_04-51-01.xlsx

#artificialintelligence

The graph represents a network of 2,067 Twitter users whose tweets in the requested range contained "#cloudcomputing", or who were replied to or mentioned in those tweets. The network was obtained from the NodeXL Graph Server on Monday, 26 October 2020 at 12:02 UTC. The requested start date was Monday, 26 October 2020 at 00:01 UTC and the maximum number of days (going backward) was 14. The maximum number of tweets collected was 7,500. The tweets in the network were tweeted over the 3-day, 9-hour, 0-minute period from Thursday, 22 October 2020 at 14:58 UTC to Sunday, 25 October 2020 at 23:58 UTC.