Re: Book Review: Programming Collective Intelligence
Rubie Mccloughan
I bought Programming Collective Intelligence years ago. My idea at that time was to use it as a basis for practical applications as I learned Python. I picked a topic that looked interesting, say genetic programming, and experimented with the example code to test my Python fluency. In the process I picked up the basics of the most common families of machine learning algorithms. Since then I've returned over and over again to Programming Collective Intelligence. But now I use the book as a refresher or introduction to different machine learning domains.
Collective intelligence (CI) is an interdisciplinary field that draws on a wide range of academic disciplines but has struggled to capitalise on cross-pollination between fields, particularly ones which do not self-identify with the collective ...
New approaches to tailoring based on collective intelligence may be able to build on the successes and lessons learned from past tailoring efforts, and may overcome the limitations inherent in current CTHC systems. Many people already encounter collective-intelligence tailoring as they interact with companies like Netflix and Amazon. These companies have developed a special class of machine learning algorithms (recommender systems) to tailor content. These systems tailor content based on collective-intelligence data (ie, data derived from the behavior of users as they interact with the system) in addition to user profiles [22-24].
Collective-intelligence data include implicit and explicit user feedback. Implicit data are derived from user actions (eg, the website view patterns of each individual accessing the system). Explicit data consist of self-reported item ratings (eg, ratings provided by users for items such as books or movies, often on a 5-star scale). However, in the health-promotion arena, patients could be asked to rate relevance, influence, or other properties of a message or product. Using these data, along with user demographic characteristics, the algorithms underlying the system generate personalized item recommendations for each user. As these systems learn more about the user, they can continually adapt to improve the recommendations.
There are also a few strategies that can be studied to overcome the limited availability of collective-intelligence data. For example, the preintervention stage of a study can be used for explicit data collection. Research is necessary to determine the minimum amount of explicit data needed to develop a reasonably functioning CTHC algorithm. Research is also needed on how to continue gathering explicit data throughout the intervention. This could be in the form of a question at the end of every message sent to the participants. Research is also warranted on how to incorporate implicit data into the algorithm as the intervention participants engage with the system (eg, visits to a website).
Hybrid systems can bridge theory-based, rule-based tailoring with the recommender empirical tailoring. While this might appear to be the best fit, it might not be feasible to develop hybrid models for all projects, given the limitations of time, content, and available collective-intelligence data. Thus, research is needed to identify the best recommender approach for an intervention and what approach would provide an advance over current rule-based approaches, make the intervention most engaging within the project constraints, and most influence the targeted behavior.
In Threadless, anyone who wants to can design a T-shirt, submit that design to a weekly contest and then rate their favorite designs. From the entries receiving the highest ratings, the company selects winning designs, puts them into production and gives prizes and royalties to the winning designers. In this way, the company harnesses the collective intelligence of a community of over 500,000 people to design and select T-shirts.
Collective Intelligence in Action is a hands-on guidebook for implementing collective-intelligence concepts using Java. It is the first Java-based book to emphasize the underlying algorithms and technical implementation of vital data gathering and mining techniques like analyzing trends, discovering relationships, and making predictions. It provides a pragmatic approach to personalization by combining content-based analysis with collaborative approaches.
Following a running example in which you harvest and use information from blogs, you learn to develop software that you can embed in your own applications. The code examples are immediately reusable and give the Java developer a working collective intelligence toolkit.
Conservationists, meanwhile, celebrate the positive impact on flora and fauna of the carefully managed mosaics of unburnt country. These allow birds and animals safe corridors between burns, and promotes plant and seed growth. The rich blend of roles is an extraordinary example of collective intelligence.
All this has come about because the Australian Government accepted that the indigenous people who managed this land for millennia should return to their historical fire and land management practices, now in partnership with modern science. The combination is a powerful example of collective intelligence benefiting climate, nature and people.
How does an AI swarm compare with findings from deep learning? A study at Stanford medical school found that groups of doctors using Swarm AI algorithms were 22 percent more accurate in making diagnoses than the most advanced deep learning algorithm that only used historical data. Clearly, having humans connected with a swarm is producing encouraging findings. DeepMind has made significant progress in creating Alpha Code to write computer code at a competitive level, where it now ranks in the top 54 percent of participants in programming competitions.5 In the spirit of collective intelligence, DeepMind is putting the dataset of problems and solutions on GitHub to spark innovation in problem-solving and code generation.
The cases of AI enabled collective intelligence illustrate a broad suite of problem-solving applications. The traditional model of internal teams focused on problem-solving has an important role to play, but even the most experienced experts will have their views placed in perspective by readily available swarm platforms.
Collective intelligence (CI) in organizational teams has been predominantly understood and explained in terms of the quality of the outcomes that the team produces. This manuscript aims to extend the understanding of CI in teams, by disentangling the core of actual collective intelligent team behavior that unfolds over time during a collaboration period. We posit that outcomes do support the presence of CI, but that collective intelligence itself resides in the interaction processes within the team. Teams behave collectively intelligent when the collective behaviors during the collaboration period are in line with the requirements of the (cognitive) tasks the team is assigned to and the (changing) environment. This perspective results in a challenging, but promising research agenda armed with new research questions that call for unraveling longitudinal fine-grained interactional processes over time. We conclude with exploring methodological considerations that assist researchers to align concept and methodology. In sum, this manuscript proposes a more direct, thorough, and nuanced understanding of collective intelligence in teams, by disentangling micro-level team behaviors over the course of a collaboration period. With this in mind, the field of CI will get a more fine-grained understanding of what really happens at what point in time: when teams behave more or less intelligently. Additionally, when we understand collectively intelligent processes in teams, we can organize targeted interventions to improve or maintain collective intelligence in teams.
As marketing professionals communicate value and manage customer relationships, they must target changing markets, and personalize offers to individual customers. With the recent adoption of large-scale, Internet-based information systems, marketing professionals now face large volumes of complex data, including detailed purchase and service transactions, social network links, click streams, blogs, comments and inquiries. While traditional marketing methodologies struggled to produce actionable insights from such information quickly, emerging collective intelligence techniques enable marketing professionals to understand and act on the observed behaviors, preferences and ideas of groups of people. Marketing professionals apply collective intelligence technology to create behavioral models and apply them for targeting and personalization. As they analyze preferences, match products to customers, discover groups of similar consumers, and construct pricing models, they generate significant competitive advantage. In this chapter, we highlight publications of interest, describe analytic processes, review techniques, and present a case study of matching products to customers.
Majority and confidence rules have been used successfully in high-stakes domains such as breast and skin cancer detection (Kurvers et al., 2015), while the seniority rule is less common (Kämmer et al., 2017). Pooling the independent judgments of small groups of diagnosticians substantially increases performance relative to average individual performance, often better than the highest performing member. The best rule often depends on the size of the group, but in general, if the decisions being pooled are unbiased, diverse, and derived independently, then the collective output will typically outperform even the best member of the group (Surowiecki, 2004). All three of these decision rules are often used in practice across a range of applied contexts, but they can lead to very different outcomes. But what about fingerprint analysis? Is it more sensible to follow the majority, the most confident, or the most senior examiner?
aa06259810