Legal document clustering with built-in topic segmentation

Lü, Qiang; Conrad, Jack G.; Al-Kofahi, Khalid; Keenan, William

doi:10.1145/2063576.2063636

Cited by 46 publications

(24 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Some individuals within the AI and Law community take performance evaluation seriously because they may be developing a commercial system that needs to be the best of its breed, not to mention to avoid litigation based on its results. For this reason, it is not uncommon to find three or four distinct tests performed on the system and documented before certification and release [4,9]. But what about the case of theoretical works within the community?, which is surely a question that will arise.…”

Section: Why Should Legal Applications Be Evaluated?mentioning

confidence: 99%

The role of evaluation in AI and law

Conrad

Zeleznikow

2015

Proceedings of the 15th International Conference on Artificial Intelligence and Law

Self Cite

View full text Add to dashboard Cite

This paper explores the presence and forms of evaluation in articles published in the journal Artificial Intelligence and Law for the ten-year period from 2005 through 2014. It represents a meta-level study of some the most significant works produced by the AI and Law community, in this case nearly 140 research articles published in the AI and Law journal. It also compares its findings to previous work conducted on evaluation appearing in the Proceedings of the International Conference on Artificial Intelligence and Law (ICAIL). In addition, the paper highlights works harnessing performance evaluation as one of their chief scientific tools and the means by which they use it. It extends the argument for why evaluation is essential in formal Artificial Intelligence and Law reports such as those in the journal. As in the case of two earlier works on the topic, it pursues answers to the questions: how good is the system, algorithm or proposal?, how reliable is the approach or technique?, and, ultimately, does the method work? The paper investigates the role of performance evaluation in scientific research reports, underscoring the argument that a performance-based 'ethic' signifies a level of maturity and scientific rigor within a community. In addition, the work examines recent publications that address the same critical issue within the broader field of Artificial Intelligence.

show abstract

Section: Why Should Legal Applications Be Evaluated?mentioning

confidence: 99%

The role of evaluation in AI and law

Conrad

Zeleznikow

2015

Proceedings of the 15th International Conference on Artificial Intelligence and Law

Self Cite

View full text Add to dashboard Cite

show abstract

“…Our work is in the field of agglomerative hierarchical clustering. Another relevant work in this field is [7] where the proposed approach consists in a two-phase clustering procedure: in a first step the information related to the instances features is exploited obtaining a "feature classification structure" that will be used by the clustering algorithm to speed-up and refine the process. With respect to these approaches, the HC f + algorithm presented in Section 4 presents two main differences.…”

Section: Related Workmentioning

confidence: 99%

“…The former is about the way the similarity values are employed. In HC f + the similarity between two ldis lead the agglomeration process while in [7] is the feature-relative similarity that leads the groups creation. The latter is about the cohesion reasons of the obtained clusters: in HC f + the features are guaranteed to be cohesive for each pair of ldis in a cluster.…”

Section: Related Workmentioning

confidence: 99%

Linked data classification

Ferrara

Genta

Montanelli

2013

Proceedings of the Joint EDBT/ICDT 2013 Workshops

View full text Add to dashboard Cite

The availability of large collections of linked data that can be accessed through public services and search endpoints requires methods and techniques for reducing the data complexity and providing high-level views of data contents defined according to users specific needs. To this end, a crucial step is the definition of data classification methods and techniques for the thematic aggregation of linked data. In this paper, we propose matching and clustering techniques specifically conceived for linked data classification, by focusing on the high level of heterogeneity of data descriptions in terms of the number and kind of their descriptive features.

show abstract

“…It is also critical that most if not all legal documents (regardless of their type) be linked to these clusters. The clusters, described in prior work (Lu et al, 2011), are meant to contain the most important case law documents on a legal topic. Yet they are also populated with other types of legal documents such as statutes, regulations, and court briefs, to ensure comprehensive coverage.…”

Section: Recommendation Via Cluster Associationmentioning

confidence: 99%

“…In short, this cluster recommendation system is enabled by the document-to-cluster association process described in Section 4. The process by which the clusters have been topically defined and populated is described in an earlier work (Lu et al, 2011).…”

Section: Introductionmentioning

confidence: 99%

Bringing Order to Legal Documents - An Issue-based Recommendation System Via Cluster Association

Lü

Conrad

2012

Proceedings of the International Conference on Knowledge Engineering and Ontology Development

View full text Add to dashboard Cite

Abstract:The task of recommending content to professionals (such as attorneys or brokers) differs greatly from the task of recommending news to casual readers. A casual reader may be satisfied with a couple of good recommendations, whereas an attorney will demand precise and comprehensive recommendations from various content sources when conducting legal research. Legal documents are intrinsically complex and multi-topical, contain carefully crafted, professional, domain specific language, and possess a broad and unevenly distributed coverage of issues. Consequently, a high quality content recommendation system for legal documents requires the ability to detect significant topics from a document and recommend high quality content accordingly. Moreover, a litigation attorney preparing for a case needs to be thoroughly familiar the principal arguments associated with various supporting opinions, but also with the secondary and tertiary arguments as well. This paper introduces an issue-based content recommendation system with a built-in topic detection/segmentation algorithm for the legal domain. The system leverages existing legal document metadata such as topical classifications, document citations, and click stream data from user behavior databases, to produce an accurate topic detection algorithm. It then links each individual topic to a comprehensive pre-defined topic (cluster) repository via an association process. A cluster labeling algorithm is designed and applied to provide a precise, meaningful label for each of the clusters in the repository, where each cluster is also populated with member documents from across different content types. This system has been applied successfully to very large collections of legal documents, O(100M), which include judicial opinions, statutes, regulations, court briefs, and analytical documents. Extensive evaluations were conducted to determine the efficiency and effectiveness of the algorithms in topic detection, cluster association, and cluster labeling. Subsequent evaluations conducted by legal domain experts have demonstrated that the quality of the resulting recommendations across different content types is close to those created by human experts.

show abstract

Legal document clustering with built-in topic segmentation

Cited by 46 publications

References 26 publications

The role of evaluation in AI and law

The role of evaluation in AI and law

Linked data classification

Bringing Order to Legal Documents - An Issue-based Recommendation System Via Cluster Association

Contact Info

Product

Resources

About