Human Computation (Synthesis Lectures on Artificial by Edith Law, Luis von Ahn

By Edith Law, Luis von Ahn

Human computation is a brand new and evolving study zone that facilities round harnessing human intelligence to resolve computational difficulties which are past the scope of latest man made Intelligence (AI) algorithms. With the expansion of the internet, human computation platforms can now leverage the talents of an exceptional variety of humans through the internet to accomplish complicated computation. there are many genres of human computation functions that exist this day. video games with a objective (e.g., the ESP online game) particularly aim on-line players who generate precious info (e.g., picture tags) whereas enjoying an relaxing online game. Crowdsourcing marketplaces (e.g., Amazon Mechanical Turk) are human computation structures that coordinate employees to accomplish initiatives in alternate for financial rewards. In id verification projects, clients practice computation for you to achieve entry to a few on-line content material; an instance is reCAPTCHA, which leverages hundreds of thousands of clients who clear up CAPTCHAs each day to right phrases in books that optical personality popularity (OCR) courses fail to acknowledge with walk in the park. This ebook is aimed toward attaining 4 pursuits: (1) defining human computation as a learn region; (2) offering a accomplished evaluate of present paintings; (3) drawing connections to a large choice of disciplines, together with AI, desktop studying, HCI, Mechanism/Market layout and Psychology, and taking pictures their precise views at the center learn questions in human computation; and (4) suggesting promising study instructions for the long run. desk of Contents: creation / Human Computation Algorithms / Aggregating Outputs / activity Routing / knowing employees and Requesters / The artwork of Asking Questions / the way forward for Human Computation

Show description

Read Online or Download Human Computation (Synthesis Lectures on Artificial Intelligence and Machine Learning) PDF

Similar intelligence & semantics books

Cambrian Intelligence: The Early History of the New AI

This publication is a suite of the "best" / so much brought up Brooks papers. primarily it covers what's thought of the center of papers that received behaviour established robotics rolling. just about all papers have seemed as magazine papers past and this is often simply a handy selection of those. For an individual engaged on cellular robotics those papers are a needs to.

Recent Advances in AI Planning: 4th European Conference on Planning, ECP'97, Toulouse, France, September 24 - 26, 1997, Proceedings

This publication constitutes the refereed court cases of the 4th ecu convention on making plans, ECP'97, held in Toulouse, France, in September 1997. The 35 revised complete papers awarded have been conscientiously reviewed and chosen from ninety submissions. the diversity of themes coated spans all facets of present man made intelligence making plans, from theoretical and foundational issues to real making plans of structures and purposes in numerous parts.

Artificial Intelligence: Its Scope and Limits

This sequence will comprise monographs and collections of experiences dedicated to the research and exploration of data, info, and knowledge­ processing structures of every kind, regardless of even if human, (other) animal, or desktop. Its scope is meant to span the complete diversity of pursuits from classical difficulties within the philosophy of brain and philosophical psycholo­ gy via matters in cognitive psychology and sociobiology (concerning the psychological services of different species) to rules relating to man made in­ telligence and to machine technological know-how.

Application of Evolutionary Algorithms for Multi-objective Optimization in VLSI and Embedded Systems

This publication describes how evolutionary algorithms (EA), together with genetic algorithms (GA) and particle swarm optimization (PSO) can be used for fixing multi-objective optimization difficulties within the region of embedded and VLSI procedure layout. Many complicated engineering optimization difficulties should be modelled as multi-objective formulations.

Additional resources for Human Computation (Synthesis Lectures on Artificial Intelligence and Machine Learning)

Sample text

With the exception of Shahaf et al. [287], there has been very little work on formalizing the notion of complexity in human computation. 1 CORRECTNESS To check for correctness of automated algorithms, specifications are commonly used—one defines the pre-conditions (the state of the variables before the execution of the program) and the expected postconditions (the state of the variables after the execution of the program), and compares the two in order to establish correctness. When computation involves humans in the loop, however, the outputs of any of the operations can be noisy, either due to mistakes or variability in judgments and opinions.

In these cases, it is difficult to derive an objective, true answer. This is because “serenity,” “pornography” and “attractiveness” are perceptual concepts that are difficult to define precisely, and their interpretation can vary greatly from culture to culture, or even from individuals to individuals. For example, mothers may be more conservative than college students in judging whether a website is pornographic or not [158], and what facial features are considered beautiful can vary depend on the cultural background of the individual annotators.

Experts can have many advantages over non-experts. In the study of expertise [63], it has been shown that experts are better at generating better, faster and more accurate solutions, detecting features and deeper structures in problems, adding domain-specific and general constraints to problems, self monitoring and judging the difficulty of the task, choosing effective strategies, actively seeking information and resources to solve problems, and retrieving domain knowledge with little cognitive effort.

Download PDF sample

Rated 4.76 of 5 – based on 23 votes