Automated Software Agents Research Articles

Insights into elections: An ensemble bot detection coverage framework applied to the 2018 U.S. midterm elections.

The participation of automated software agents known as social bots within online social network (OSN) engagements continues to grow at an immense pace. Choruses of concern speculate as to the impact social bots have within online communications as evidence shows that an increasing number of individuals are turning to OSNs as a primary source for information. This automated interaction proliferation within OSNs has led to the emergence of social bot detection efforts to better understand the extent and behavior of social bots. While rapidly evolving and continually improving, current social bot detection efforts are quite varied in their design and performance characteristics. Therefore, social bot research efforts that rely upon only a single bot detection source will produce very limited results. Our study expands beyond the limitation of current social bot detection research by introducing an ensemble bot detection coverage framework that harnesses the power of multiple detection sources to detect a wider variety of bots within a given OSN corpus of Twitter data. To test this framework, we focused on identifying social bot activity within OSN interactions taking place on Twitter related to the 2018 U.S. Midterm Election by using three available bot detection sources. This approach clearly showed that minimal overlap existed between the bot accounts detected within the same tweet corpus. Our findings suggest that social bot research efforts must incorporate multiple detection sources to account for the variety of social bots operating in OSNs, while incorporating improved or new detection methods to keep pace with the constant evolution of bot complexity.

Bots fired: examining social bot evidence in online mass shooting conversations

Mass shootings, like other extreme events, have long garnered public curiosity and, in turn, significant media coverage. The media framing, or topic focus, of mass shooting events typically evolves over time from details of the actual shooting to discussions of potential policy changes (e.g., gun control, mental health). Such media coverage has been historically provided through traditional media sources such as print, television, and radio, but the advent of online social networks (OSNs) has introduced a new platform for accessing, producing, and distributing information about such extreme events. The ease and convenience of OSN usage for information within society’s larger growing reliance upon digital technologies introduces potential unforeseen risks. Social bots, or automated software agents, are one such risk, as they can serve to amplify or distort potential narratives associated with extreme events such as mass shootings. In this paper, we seek to determine the prevalence and relative importance of social bots participating in OSN conversations following mass shooting events using an ensemble of quantitative techniques. Specifically, we examine a corpus of more than 46 million tweets produced by 11.7 million unique Twitter accounts within OSN conversations discussing four major mass shooting events: the 2017 Las Vegas concert shooting, the 2017 Sutherland Springs church chooting, the 2018 Parkland School Shooting and the 2018 Santa Fe school shooting. This study’s results show that social bots participate in and contribute to online mass shooting conversations in a manner that is distinguishable from human contributions. Furthermore, while social bots accounted for fewer than 1% of total corpus user contributors, social network analysis centrality measures identified many bots with significant prominence in the conversation networks, densely occupying many of the highest eigenvector and out-degree centrality measure rankings, to include 82% of the top-100 eigenvector values of the Las Vegas retweet network.

Operationalizing Conflict and Cooperation between Automated Software Agents in Wikipedia

This paper replicates, extends, and refutes conclusions made in a study published in PLoS ONE ("Even Good Bots Fight"), which claimed to identify substantial levels of conflict between automated software agents (or bots) in Wikipedia using purely quantitative methods. By applying an integrative mixed-methods approach drawing on trace ethnography, we place these alleged cases of bot-bot conflict into context and arrive at a better understanding of these interactions. We found that overwhelmingly, the interactions previously characterized as problematic instances of conflict are typically better characterized as routine, productive, even collaborative work. These results challenge past work and show the importance of qualitative/quantitative collaboration. In our paper, we present quantitative metrics and qualitative heuristics for operationalizing bot-bot conflict. We give thick descriptions of kinds of events that present as bot-bot reverts, helping distinguish conflict from non-conflict. We computationally classify these kinds of events through patterns in edit summaries. By interpreting found/trace data in the socio-technical contexts in which people give that data meaning, we gain more from quantitative measurements, drawing deeper understandings about the governance of algorithmic systems in Wikipedia. We have also released our data collection, processing, and analysis pipeline, to facilitate computational reproducibility of our findings and to help other researchers interested in conducting similar mixed-method scholarship in other platforms and contexts.

Bots, bespoke, code and the materiality of software platforms

This article introduces and discusses the role of bespoke code in Wikipedia, which is code that runs alongside a platform or system, rather than being integrated into server-side codebases by individuals with privileged access to the server. Bespoke code complicates the common metaphors of platforms and sovereignty that we typically use to discuss the governance and regulation of software systems through code. Specifically, the work of automated software agents (bots) in the operation and administration of Wikipedia is examined, with a focus on the materiality of code. As bots extend and modify the functionality of sites like Wikipedia, but must be continuously operated on computers that are independent from the servers hosting the site, they involve alternative relations of power and code. Instead of taking for granted the pre-existing stability of Wikipedia as a platform, bots and other bespoke code require that we examine not only the software code itself, but also the concrete, historically contingent material conditions under which this code is run. To this end, this article weaves a series of autobiographical vignettes about the author's experiences as a bot developer alongside more traditional academic discourse.

Ethical Approaches to Robotic Data Gathering in Academic Research

Internet researchers increasingly have at their disposal of an array of automated software agents, or “bots,” which can rapidly and efficiently retrieve a variety of economic and technical data from publicly accessible web sites. While these automated tools greatly facilitate the retrieval and analysis of data for academic research, they may pose ethical problems for Internet researchers. Specifically, automated software bots place some load on servers being accessed, possibly in contradiction to the expected use of such servers, and possibly in violation of the legal prerogatives of web site owners. Determining how and when to access such web sites, and whether to seek the consent of web site owners for retrieval of publicly accessible data presents an apparent conflict between general principles of information policy and the emerging legal precedent regarding trespass to computers. This conflict may be characterized as pitting utilitarian considerations against deontological considerations in a fashion reminiscent of previous debates over informed consent in on-line research. In this paper, we examine both utilitarian and deontological characterizations of the ethical obligations of researchers employing automated data retrieval bots, and argue that the contrasts between the two approaches do not necessarily result in conflict. Instead, we argue that the tension within the relevant practices indicates the need for a “meta-choice” between utilitarian and deontological considerations. We further suggest certain factors that may differentiate such a “metaethical” choice in the context of automated data retrieval from the “meta-ethical” choice presented in previously identified contexts of human subjects research or of web browser technology design. In the end, we argue that by analyzing the ethical issues in terms of the contrast between utilitarian and deontological ethics, it is possible to resolve some of the ethical dilemmas regarding automated data retrieval in fruitful and cogent ways.

Economic assessment of externalities for interactive audio media anti-SPIT protection of internet services

Spam over internet telephony (SPIT) refers to all unsolicited and massive scale attempts to establish voice communication with oblivious users of voice over internet protocol (VoIP) services. SPIT exhibits a significant increase over the last years, thus developing into a serious threat with adverse impact and costs for the business economy. An audio completely automated public Turing test to tell computers and human apart (CAPTCHA) has been introduced as a means to distinguish automated software agents (bots) from human. CAPTCHA has been proposed as a security measure against SPIT. In this paper, we lay the principles for an adequate understanding of the SPAM-related economic models, as well as their analogies to the SPIT phenomenon, so as to weigh the benefits of audio CAPTCHA protection against the incurred costs. Our approach is based on the economic assessment of externalities, i.e., the economic impact associated with the SPIT side effects on the everyday life.

CMieux: Adaptive strategies for competitive supply chain trading

Supply chains are a central element of today’s global economy. Existing management practices consist primarily of static interactions between established partners. Global competition, shorter product life cycles and the emergence of Internet-mediated business solutions create an incentive for exploring more dynamic supply chain practices. The supply chain trading agent competition (TAC SCM) was designed to explore approaches to dynamic supply chain trading between automated software agents. TAC SCM pits trading agents developed by teams from around the world against one another. Each agent is responsible for running the procurement, planning and bidding operations of a PC assembly company, while competing with others for both customer orders and supplies under varying market conditions. This paper presents Carnegie Mellon University’s 2005 TAC SCM entry, the CMieux supply chain trading agent. CMieux implements a novel approach for coordinating supply chain bidding, procurement and planning, with an emphasis on the ability to rapidly adapt to changing market conditions. We present empirical results based on 200 games involving agents entered by 25 different teams during what can be seen as the most competitive phase of the 2005 tournament. Not only did CMieux perform among the top five agents, it significantly outperformed these agents in procurement while matching their bidding performance. We also simulated 40 games against the best publicly available agent binaries. Our results show CMieux has significantly better average overall performance than any of these agents.

Supporting group interaction among humans and autonomous agents

An important aspect of interaction among groups of humans and software agents is supporting collaboration among these heterogeneous agents while they operate remotely and communicate asynchronously. We are developing an architecture that supports multiple humans interacting with multiple automated control agents in such a manner. We are evaluating this architecture with a group consisting of the crew of a space-based vehicle and the automated software agents controlling the vehicle systems. Such agent interaction is modelled as a loosely co-ordinated group because this model minimizes agent commitment to group goals and constraints while addressing a significant portion of crew and control agent group behaviours. In this paper, we give background on human interaction with space-based automation. We identify related research in multi-agent autonomous architectures and single agent human-computer interaction systems, we describe our architecture design for human-software agent groups and we identify research issues in loosely co-ordinated human-software groups.

Internet Comparative Mapping Resources.

Databases accessible over the Internet have become an essential tool in the study of genome organization. Recent growth in the field of genomics has resulted in the study of an increasing number of organisms in addition to a flood of data from the more intensively studied organisms. Although this rapid growth in the quantity of data has greatly increased the likelihood that data useful to an individual researcher will be available, the job of finding that data in the traditional journal publication system is significantly more difficult. To increase the accessibility of these data, several databases have been established and made available via the Internet. This availability allows the rapid access of current, in-depth data not practical for journal publications, with the additional benefit of enabling searchers to extract the data based on their own criteria. Genome databases have been established on an organism-specific basis, usually with a proprietary database structure and delivery system. The largest of these databases are the Human Genome Organization's genome database at Oak Ridge National Laboratory (Oak Ridge, Tennessee) and the Mouse Genome Database at The Jackson Laboratory (Bar Harbor, Maine). Many of the species groups are now migrating to a common database platform; and several species groups including those working on pig, cattle, sheep, horse, chicken, and cat are now using the ARKdb platform developed and run at the Roslin Institute in Edinburgh, Scotland. This usage of a common system holds great promise for the integration and cross-referencing of data for these species. At the time of this writing, the status and direction of the largest genome database—the human genome database—are uncertain due to the termination of funding for its development. The database will be hosted in its current form at Oak Ridge National Laboratory with no additional development, representing a shift in the focus from geneand clone-based information to a model of annotated sequence. This shift is likely to present some difficulties for comparative mapping in the short term. Comparative mapping data are integrated into many of the species-specific databases in several ways. The basic type of comparative linkage comprises shared symbols and names and includes the use of symbols from other species as an alias. This type of linkage has many flaws, including inflexibility with regard to our evolving understanding of gene homologies and sensitivity to nomenclature changes. The more advanced comparative information is obtained from direct links by unique identifiers to entries in other databases and in some cases dedicated detailed comparative databases. The most comprehensive comparative database available as of October 1998 is the Mouse Genome Database (mammalian homology database), which lists homologies in a broad range of species as given in published journal articles (with references) and includes a powerful set of search criteria in a forms interface with tabular or graphical output. The main limitations of all current comparative mapping resources are the manual nature of establishing homology links and the typical absence of database entries that indicate the evidence for the homology. The next wave of genome database innovation will be the development of automated software agents, which will explore databases and create their own database of links between the different species databases based on a defined set of homology criteria. One of these agent systems, under development at the time of this writing at the Roslin Institute and the University of Edinburgh, will hopefully result in the efficient creation of current, robust, and well-defined homology links between species databases. Although comparative gene mapping databases are continuing to evolve, Internet-based resources have become an entrenched and vital tool for the comparative mapper. Use of this tool can only grow in importance as the data sets involved continue to expand and as more sophisticated tools are developed for their exploitation.