Character-Based Models for Adversarial Phone Extraction: Preventing Human Sex Trafficking

Chambers, Nathanael; Forman, Timothy; Griswold, Catherine; Khastgir, Yogaish; Steckler, Stephen

doi:10.18653/v1/d19-5507

Cited by 4 publications

(7 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We use the term adult service providers (ASP) to refer to a worker in the sex industry (also referred to as the sex trade) that offers sexual services. Given that an ASP is seeking paying clients and that clients must have a method by which to contact them, linking ads which use the same phone number is logical and has proved to be an effective strategy (Chambers et al 2019, Dubrawski et al 2015, Ibanez and Gazan 2016a. Over time, phone numbers have become inexpensive, that is, easy to obtain and change, making them a less reliable proxy for an individual.…”

Section: Human Trafficking and Online Sex Advertisementsmentioning

confidence: 99%

See 1 more Smart Citation

Cracking Sex Trafficking: Data Analysis, Pattern Recognition, and Path Prediction

Keskin

Bott

Freeman

2021

Production and Operations Management

View full text Add to dashboard Cite

Human trafficking, the exploitation of humans for monetary gain or benefit, is a widespread humanitarian issue that is typically sub‐classified into labor and sex trafficking. In the last decade, sex traffickers have used online classified advertisements to advertise sexual services. Although these advertisements are visible to the general public and law enforcement, the volume of ads, the frequency with which their posting locale changes, and the use of obfuscation tactics make it difficult for law enforcement agencies to react. Existing products for law enforcement focus on identifying, tracking, and correlating individual activity by performing deep searches for specific information against a database of historical posts. While this deep search capability is useful for investigating specific cases, it overlooks higher‐level patterns that exist in ads. Using a website that has been linked to several sex trafficking‐related arrests, we demonstrate a framework for harvesting, linking, and detecting these patterns in a dataset comprised of more than 10 million advertisements targeting US cities. Our framework combines information systems and operations research concepts to identify groups of posts based on text, phone numbers, and pictures; determine circuits associated with post groups, and predict future movements using four different methods. Our description of the framework and comparison of the grouping and prediction methods provide insights that can assist law enforcement agencies to combat individuals/organizations involved in illicit sexual activities, including sex trafficking, proactively. Also, this demonstration provides researchers interested in developing advanced interdiction models targeting illicit sexual activities with a clear picture regarding available data formats.

show abstract

Section: Human Trafficking and Online Sex Advertisementsmentioning

confidence: 99%

“…Given that an ASP is seeking paying clients and that clients must have a method by which to contact them, linking ads which use the same phone number is logical and has proved to be an effective strategy (Chambers et al. 2019, Dubrawski et al. 2015, Ibanez and Gazan 2016a).…”

Section: Introductionmentioning

confidence: 99%

Cracking Sex Trafficking: Data Analysis, Pattern Recognition, and Path Prediction

Keskin

Bott

Freeman

2021

Production and Operations Management

View full text Add to dashboard Cite

show abstract

“…However, these indicators can only be studied within a cluster of ads linked with individual vendor accounts. Previous work by Chambers et al (2019) proposed using neural networks to extract phone numbers to connect these escort ads. Nonetheless, our research demonstrates that only 37% of the ads in our dataset contained phone numbers.…”

Section: Related Researchmentioning

confidence: 99%

“…Figure 2(A) and figure 2(B) show that most ads in our dataset (approximately 99%) have a sentence length below 512 tokens and 2,000 characters. To generate ground truth, i.e., vendor labels, we employ the TJBatchExtractor Nagpal et al (2017b) and CNN-LSTM-CRF classifier Chambers et al (2019) to extract phone numbers from the ads. Subsequently, we utilize NetworkX Hagberg et al (2008) to create vendor communities based on these phone numbers.…”

Section: Datasetmentioning

confidence: 99%

IDTraffickers: An Authorship Attribution Dataset to link and connect Potential Human-Trafficking Operations on Text Escort Advertisements

Saxena,

Ashpole,

van Dijck

et al. 2023

Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing

View full text Add to dashboard Cite

Human trafficking (HT) is a pervasive global issue affecting vulnerable individuals, violating their fundamental human rights. Investigations reveal that many HT cases are associated with online advertisements (ads), particularly in escort markets. Consequently, identifying and connecting HT vendors has become increasingly challenging for Law Enforcement Agencies (LEAs). To address this issue, we introduce IDTraffickers, an extensive dataset consisting of 87,595 text ads and 5,244 vendor labels to enable the verification and identification of potential HT vendors on online escort markets. To establish a benchmark for authorship identification, we train a DeCLUTR-small model, achieving a macro-F1 score of 0.8656 in a closed-set classification environment. Next, we leverage the style representations extracted from the trained classifier to conduct authorship verification, resulting in a mean r-precision score of 0.8852 in an open-set ranking environment. Finally, to encourage further research and ensure responsible data sharing, we plan to release IDTraffickers for the authorship attribution task to researchers under specific conditions, considering the sensitive nature of the data. We believe that the availability of our dataset and benchmarks will empower future researchers to utilize our findings, thereby facilitating the effective linkage of escort ads and the development of more robust approaches for identifying HT indicators 1 .

show abstract

“…• TJBatch -The state-of-the-art named entity extractor (Dubrawski et al, 2015;Chambers et al, 2019) in the human trafficking domain. This method extracts words from a dictionary and is based on manually designed regex rules.…”

Section: Introductionmentioning

confidence: 99%

Extracting Person Names from User Generated Text: Named-Entity Recognition for Combating Human Trafficking

Li¹,

Nair²,

Pelrine³

et al. 2022

Findings of the Association for Computational Linguistics: ACL 2022

View full text Add to dashboard Cite

Online escort advertisement websites are widely used for advertising victims of human trafficking. Domain experts agree that advertising multiple people in the same ad is a strong indicator of trafficking. Thus, extracting person names from the text of these ads can provide valuable clues for further analysis. However, Named-Entity Recognition (NER) on escort ads is challenging because the text can be noisy, colloquial and often lacking proper grammar and punctuation. Most existing state-of-the-art NER models fail to demonstrate satisfactory performance in this task. In this paper, we propose NEAT (Name Extraction Against Trafficking) for extracting person names. It effectively combines classic rule-based and dictionary extractors with a contextualized language model to capture ambiguous names (e.g penny, hazel) and adapts to adversarial changes in the text by expanding its dictionary. NEAT shows 19% improvement on average in the F1 classification score for name extraction compared to previous state-of-the-art in two domain-specific datasets.

show abstract

Character-Based Models for Adversarial Phone Extraction: Preventing Human Sex Trafficking

Cited by 4 publications

References 17 publications

Cracking Sex Trafficking: Data Analysis, Pattern Recognition, and Path Prediction

Cracking Sex Trafficking: Data Analysis, Pattern Recognition, and Path Prediction

IDTraffickers: An Authorship Attribution Dataset to link and connect Potential Human-Trafficking Operations on Text Escort Advertisements

Extracting Person Names from User Generated Text: Named-Entity Recognition for Combating Human Trafficking

Contact Info

Product

Resources

About