INTRODUCTION
Naturally, the main aim of legal information retrieval systems is to find relevant documents in relation to the search query as defined by the user. Major litigation consists of three stages – pre-trial, trial and appeal; but information sought by the litigant is similar in nature at each stage. The pre-trial stage involves intensive document discovery while in the trial stage the main focus is on providing fast retrieval, and tracking, of various documents that have been introduced in evidence. At the appeal stage the main focus is on trial and case law. In the digital era various search techniques are used to explore the legal databases and these are different from other conventional search techniques.
Raj Kumar Bhardwaj
SEARCHING LEGAL DATABASES: SOME THEORY AND A LITERATURE REVIEW
Yuan (1997)Footnote 1 monitored the LexisNexis quick law searches of a group of law students over a period of one year. Yuan examined several aspects of their searching behaviour, including the increase of their command and feature repertoires, their change in language usage, the increase of search speed and the change of learning approaches. Yuan discovered that experience did not result in either searchers making fewer errors, or being able to recover from more errors. Yuan also found that although participants with higher levels of quick law experience used a greater variety of commands and features than those with lower levels of experience, some commands remained rarely or never used.
Oulanov and Pajarillo (2003)Footnote 2 also conducted a study on perceptions of LexisNexis, this time using structured questionnaires that were issued to eight academic librarians at Queens Borough Community College in New York City. Although the authors did not include a copy of the questionnaire used, the tables that were included reveal that the questionnaire was predominantly based on a five point Likert grading scale aimed at uncovering the librarians' perceptions on three aspects of the resource; its ‘retrieval features’; its ‘effectiveness’ and other usability-related aspects which the authors rather ambiguously categorise under the heading ‘user effort perception criterion’.
Komlodi and Soergel (2002)Footnote 3 also focused on information use and re-use, specifically on legal information seekers and the use of their memory and externally recorded search histories to inform their later searches. Komlodi and Soergel found, like Kuhlthau and Tama (2001)Footnote 4 and Blomberg et al. (1996)Footnote 5, that during the legal research process, law students not only needed to consult electronic legal resources, but also needed to return to their personal research files. Komlodi and Soergel developed a set of search-history-based user interface tools to support the recording, categorisation and annotation of search results. These were achieved by the system keeping track of user actions and results, and using this expanded history to encourage easier information re-use and future search tasks. The work led to a form of search histories being incorporated into the Westlaw electronic legal resource.
Marshall et al. (2001)Footnote 6 conducted a study and identified the continued importance and authority of books in students’ legal research process. The study found that many of the users' information-seeking strategies followed links rather than conducting explicit searches and highlighted the use of electronic resources for case evaluation. Marshall et al. stated in their study that students began their moot court research by identifying case law and described this as a ‘launching pad’ or ‘looking for a thread to pull!’ The students then continued to use citations as a point of departure, either as obvious links to a precedent if they came across the citation several times or as a way of determining whether the cases were still ‘good law’.
Essential Search techniques
In legal text searching a Natural Language Query is one that is expressed using normal conversational syntax. The benefits of using natural language queries are that it can take less time than traditional Boolean queries.
Fuzzy words can be used in the feedback searching process to gain added insight into the nature of databases. The fuzzy operator scans the database for words with a spelling similar to that of a query term. It can recognise incorrect spelling and instances where scanned materials have been incorrectly integrated during the Optical Character Recognition (OCR) process. Fuzzy matching looks for words that are similarly spelt in the database. Operators can be helpful in finding word variants or misspellings in the database, the result of typographically mistakes or errors during the process of OCR.
LEGAL SEARCH STRATEGIES AND RETRIEVAL SYSTEMS
The components of legal text retrieval systems can be categorised in the following way:
a. Text-file Search;
b. Discarded stop words;
c. Index to search file;
d. Interest retrieval;
e. Fact Retrieval;
f. Coverage of collection;
g. Speed of retrieval;
h. Recall of system;
i. Precision of system;
j. User friendliness.
The six qualities of information retrieval systems, as listed by Cyril Cleverdon in his evaluation of information retrieval,Footnote 7 are:
Legal search strategies
Coverage of Collection
Typically, legal databases contain case law of various High Courts and Supreme Courts; and articles published in journals together with speeches of eminent personalities from the legal fraternity.
Recall
Recall is the ratio of the number of relevant records retrieved compared to the total number of relevant records in the database.
Precision
Precision is the ratio of the number of relevant records retrieved compared to the total number of irrelevant and relevant records retrieved.
Fall-out
The term, fall-out, refers to all the ‘junk’ received from the search that is irrelevant. If 100 documents are retrieved and 20 are relevant, then fallout is 80%. Fallout becomes a larger issue as the size of the database grows.
Presentation of results
In most cases the results that are displayed are formatted as a vertical list of retrieved documents. An item in the displayed list consists of the title of the document and a set of important metadata such as the party name, the advocate, the name of the judge and the date of decision, along with a headnote of the case.
Time and effort involved to obtain answers
The time and effort involved in the process to retrieve the relevant records should be minimal and it should be user-friendly so that users can access the system without any constraints (Kumar et al. 2005).Footnote 8
PERFORMANCE OF TEXT RETRIEVAL
There is a significant difference between interest and fact retrieval. Lawyers use fact retrieval for mainly three purposes; i) to find dates, either within the documents or as part of the identification of a document ii) lawyers search for document identifications, names, cases, citation of statutes iii) to look for party names, name of advocates, name of judges, and so forth.
In fact, the retrieval of documents is designed so that the date of the case and the text of the case itself has the same weightage. Failures to retrieve documents are interpreted as precious information provided that there are no errors in the material; denoting that there is no citation to the case in question. Interest retrieval is characterised by a difference in the assessment of relevance as it is not possible to specify which relevant text a particular document contains.
ELECTRONIC LEGAL DATABASES
Naturally, in India as eleswehere, electronic databases play a major role in libraries. In order to fulfill the individual and varied needs of their users, law libraries subscribe to a host of electronic services. Indian law libraries subscribe to, and make significant use of, the following databases – HeinOnline, Westlaw, Manupatra, Grand Jurix, SCC online, A.I.R Online, and Lexis. Two services, of particular value, among Indian law libraries are Westlaw, which gives access to more than 4,500 international online journals, and Hein Online, which provide access to more than 1,500 journals.
The databases listed above each provide a range of facilities by which numerous operators can be used to retrieve relevant case law. A review of these services and their retrieval systems is provided below.
COMMERCIAL LEGAL INFORMATION SYSTEMS: A SYNOPSIS
SCC Online Case Finder
The SCC (Supreme Court Cases) Online Case Finder is a proprietary product of EBC Publishing Pvt. Ltd based in Lucknow, INDIA. It has an installation of two CD-ROMs (a) Typical Mode (b) Minimal Mode (Figure 2).
SCC Online Case Finder
The ‘Search Results’ dialog box contains separate lists of the case-notes found for each database (modes 1 and 2) that is searched. They are alphabetically arranged by the topic/statute headings under which they occurred. For judgments, the sorting order is by the date of the decision of the case and the first few case-note headings are displayed on the screen. Typically, the total number of case-notes found can often exceed the number of case-notes that can be displayed in the visible portion of the list.
All India Reporter(AIR) on CD-ROM
This database offers full text judgments of the Supreme Courts of India and all High Courts. It gives the headnotes, citation search, free text search, search by party name, the name of the judge and statutes name, and so on. The advanced query also gives complete access to the Folio Views Query Syntax. This syntax helps to focus and refine searches through the use of Boolean operators, wildcards, proximity operators and scope limitations. The Query option gives more information about performing simple searches (Figure 3).
SCC Online Case Finder – search screen
LexisNexis India
LexisNexis is the product division of Reed Elsevier India Pvt. Ltd. It covers all Supreme Court Cases (since inception), updated legal acts and articles from selected legal journals. There are also new editions of commentaries by eminent legal authors. It has a helpful my book shelf facility and, in terms of search opportunities, users can search all the resources from the home page. The way that the results are clustered empowers the user with multi-faceted hits for each result. Controlled vocabulary is used in the indexing of the database. The taxonomy refers to the topic level of classification of chapters (Figure 4).
LexisNexis India
Legal Pundits
This legal database was created by Legal Pundit International Service Pvt. Ltd. with the aim of delivering a gamut of legal information services to individuals and enterprises. It empowers users with both a general search and a case law search facility. The database covers the cases from the Supreme Court, various High Courts, APTEL, AAR, CAT, Company Law Board, DRAT, DRT, CERC, IPAB, ITAT, NCDRC, Privy Council, SCDRC, SEBI (SAT), STT, TDSAT, Trademark and can be searched by various subjects areasFootnote 9. However, the general search option allows the search to be expanded to commentaries and analysis, notifications, forms and procedures, circulars, rules, guidelines, schemes, drafts, bare acts, trade notices, press notes, regulations, policies, and so forth (Figure 5).
Legal Pundits.com
Searches can be saved, draft searches can be viewed and the case law search archive can be checked.
Chawla Law Finder
Chawla Law Finder is an efficient, time saving and an economical case search engine designed and developed by Chawla Publication Pvt. Ltd. The case finder contains five databases; (a) Recent Criminal Reports (b) Recent Civil Reports (c) Service Cases Today (d) Recent Control Reporter (e) Dishonour of Cheques Total Cases.
Judgments can be searched with the court name, judge name, decision date, petitioner, respondents, advocate name, head-note, and case reference order and result. This case finder gives the user advanced search features such as feedback search, concept searching and fuzzy words.
Grandjurix
Grandjurix is available in CD-ROM format as well as an online product. The electronic version is called e-Jurix, a product of Spectrum Business Support Ltd which was established in 1988. It covers 250,000 full text judgments and covers all Supreme Court, High Court and Tribunal Decisions reported to date. eJurix provides search facilities by full text, subject, section-act, title, keywords phrases, statutes referred, quorum of judges, name of the court, date of decisions, and equivalent citations. It also includes the full text of judgments of appropriate quasi-judicial bodies, High Courts and the Supreme Court of India, from 1950, as well as basic Information – acts rules and regulations, and notifications & circulars issued by the Law Enforcement agencies. eJurix has a personalisation facility which allows the user to store expertise and save the search terms in the database (Figure 6).
eJurix
The My preferences option in the side bar assists the user to set the search settings; it helps users to search the title, the contents and the annotations. Proximity options extend the facility to search anywhere – i.e. a phrase search, search within a sentence, search within a paragraph and to search with up to 40 words. Acts, rules and notifications can be searched as well as be browsed alphabetically. The advanced search feature makes it more robust. Users can select court name, judge name and year of judgment. The volume number and page can also be searched (Figure 7).
eJurix – Case law searching
Manupatra
Online Manupatra.com was launched in the year 2001 with a wide variety of content – i.e. commentaries, digest, editorial enhancements, treaties, case laws and more. It empowers its user with legislative and procedural information. It covers Supreme Court Cases from 1950 to date, cases from the high courts, tribunals and commissions. It also contains acts, notifications and circulars, forms, draft agreements, WTO, materials relating to arbitration, cyber laws, intellectual property law, labour and employment law, human rights, environmental law, and media and communication laws.
In addition, Manupatra gives access to e-books, electronic articles and has an international aspect to the database too. It has a facility for equalling citations of multiple print journals. Efficient hyperlinks, to referred judgments, assist the legal researcher. Overruled and reversed judgments can also be identified in Manuptra. Each section within the judgments is hyperlinked so that researchers can access the bare acts instantly. Searches can also be achieved with the citation, i.e. volume no, year, page no., etc. More than 1,100 bare acts are regularly added which includes various amendments and repealed acts (Figure 8).
Manupatra
THE OPERATORS USED IN LEGAL TEXT RETRIEVAL: A REAPPRAISAL
Single Character wildcard operator
In search retrieval, the single character wildcard, i.e. the question mark, is used to represent a single variable character in a given query; eg. the syntax would be Str??ing for string; Exclu???? will find excluding, exclusive.
In single character wildcard searching, another option is matching one, or more, characters; employing the optional wildcard which represents one, or more, variable characters. Egs. Colo$r, V$TOL, Electron$$$. The first query will find the terms like color and colour. The second query will search for terms like VTOL, VSTOL. The third query would search for terms like electron, electrons, electronic and electronics.
Matching a character string; the asterisk represents a variable string of one or more characters in a query. The string wildcard is equivalent to an infinite series of adjacent optional character wildcards. Examples: Medic*, m*n. The first search term will search for media, medical, medicine, medicate, medically, medication, and second will search man, men, mean, moon and moron!
The Fuzzy search operator
The fuzzy search operator is used to retrieve records that contain words with spellings similar to that of a particular query term. That part of a query term will be matched exactly as determined by the placement of the operator within the word. The character string that precedes the operator will be anchored eg. Rajn ~ h, Corr ~ o. The first query will retrieve record that contains Rajnish, or alternate spellings of the word Rajneesh and second query will retrieve records that contain corrode, or similarly spelled words like corrosive and corrosion.
Concept Operator
The concept search operator retrieves results on the basis of concept and many results will be retrieved that do not having the occurrences of the original query word.
Near Operator
This operator searches for word pairs in which the second term occurs within the specified number of words before, or after, the first. It is a bidirectional proximity operator and works across boundaries; you cannot use it to search for a word pair in which the words occupy separate places within a record. This is a search operator that is evident in the Chawla Law Finder database.
Boolean Operator
The commonly used And, Or, Not operators are often used to define the relationship between a party name, judge name, case number, court name, keywords or group. With the help of these operators, combinations of elements can be searched.
Proximity operator
This operator is used to search for word pairs in which the second term of the pair occurs within a specified number of words after the first. This operators does not work across field boundaries and it cannot be used to search for word pairs in which the word occupys separate fields within the record. Proximity operators can be further classified as the (a) Order Proximity search and (b) Unordered Proximity search. Order Proximity is used to specify which term must appear within a given range to count as a hit. This is used with a forward slash. Terms in an ordered proximity search must be enclosed in quotes. As a side note, a phrase search is, basically, an ordered proximity search with a proximity equal to the number of terms in the phrase. e.g. “Court constitution”/5. This means find records which contains ‘court’ and ‘constitution’ in that order, within a 5 words range. While unordered proximity is to specify a set of terms which must appear within a given range in any order. The unordered proximity operator is the @ symbol. Terms in an unordered proximity search word must be enclosed in quotes.
Adjacency Operator
This operator is equivalent to the proximity operator with a defined range of one word. Certain punctuation – hyphens, apostrophes, commas and periods- function as adjacency operators when they appear in the middle of character string. This operator does not work across field boundaries and this is unidirectional from left to right.
Same/n operator
This operator is a search query term that is within n paragraphs of another query term. A law finder will automatically search for documents in which query terms on either side of the same operator appear in the same paragraph. The same operator is bidirectional, eg. poison same/3 Chandigarh.
Not same operator
The Not same operator searches for query terms that do not appear in the same paragraph as the second query term. In addition this operator is unidirectional, eg. poison not same dowry. The query will retrieve documents in which poison does not fall in the same paragraphs as dowry.
At least /in operator
This operator is used to search records that contain occurrences of the query expression that immediately follows the at least operator. Eg. at least /10 soldiers will search at least 10 occurrences of the word, soldiers.
Exact Match
The exact match search for an individual query uses a sign to force a search for the exact word only. Eg. “India”. This query will search only India and not Indian or Indiana.
Exact phrase operator
The exact phrase operator – the single quotation mark –combines characteristics of the adjacency and exact match operators; with this operator we can search exact matches of a phrase such as; “code of criminal procedure.”
Equivalence operator
This operator is used to search records containing the specified alphanumeric value in the specified field (or series of fields specified by commas), eg.; Judge = Gaur and Gaur = Judge; both options search for judgments containing Gaur in the Judge field.
Range Operator
With this operator, we can search for records in which a specified field contains the specified value. The syntax of this operator is an angle bracket greater (>) than and less than equal (<) bracket followed by the equal sign (>=, <=). Example: Decision date <= 1990, Mahesh < Petitioner < Ramesh. In the above example, the first query will retrieve record in which value of 1990, or lower occur in the decision date field. The second will retrieve the record in which the name field contains values alphabetically between Mahesh and Ramesh.
Query level field restriction operator
This operator searches for records in which the query terms preceding it appear in the field (or series of fields separated by commas) that follows it.
Legal Databases and their Search Features
The table below offers a comparison of the operators and search features (as described in the text above) between various key legal databases used by researchers and scholars in relation to Indian law.
Comparative study of various search operators and search features used in legal databases
CASE STUDY OF FACULTY OF LAW, UNIVERSITY OF DELHI, DELHI
Objectives of the Study:
The following objectives were set up for the Study:
(a) to study the awareness, use and the method of using the legal information databases among library users;
(b) to study the frequency and purpose of using legal information databases;
(c) to identify the search method used in searching legal information databases;
(d) identify the method of training to use the legal information systems in a training programme;
(e) to rate the quality of legal information databases.
Methodology and Analysis of Data:
In order to know the true usage of electronic information sources, a survey was conducted in 2011 and a questionnaire was prepared to collect the data. A total of 139 questionnaires were distributed and 66.18% completed questionnaires were received.
Awareness about Legal Information Databases
Table 1 summarises the awareness of electronic resources and shows that 66% of the LLB respondents were aware of the electronic databases, whereas only 34% respondents indicated they were unaware. However, LLM students (95%) were more aware; only 5% stated they were unaware of electronic legal databases. Overall, 78.89% of the respondents were aware of legal information database; however 21.11% of the respondents were unaware of the legal information databases.
Awareness of Legal Information Database
Note: n = 90
Frequency of Use
Users were asked to mention the frequency of the usage of electronic resources with four options-frequently, occasionally, sometime and rarely. 30% of the LLB respondents indicated they frequently used the electronic resources. 40% respondents replied that they occasionally used these and 20% stated sometime, However 6% respondents rarely used the electronic resources while 4% respondents stated that they never used the legal information databases. The table shows the slight difference in frequency of usage of Legal Information Database between LLB and LLM students.
Frequency of using online Legal Information Database by LLB and LLM students
Note: n = 92
Purpose of Using Legal Information Databases
Within the questionnaire respondents were asked about the purpose of using electronic resources, using open ended questions, six reasons were enlisted with more than one option of choice. 21.34% of the respondents used legal information databases in research and development activities with 33.70% using them for projects and assignments. 33.70% respondents were in support of case law searching. As far as the purpose of using e-resources in different disciplines was concerned, 11.23 % of law students used them for study and update purpose.
Purpose of using Legal Information Database by LLB and LLM students
Note: n = 89
Search and Retrieval
The data depicted below in Table 4 explains that respondents preferred boolean operators in their search and retrieval. While 91.30% of the respondents preferred the simple search, interestingly, 84.78% of the respondents opted for the boolean search method. None of the respondents selected the command level search and retrieval. Where the proximity search was concerned only 23.91% of the respondents stated they used this method for searching and retrieval and 76.08% mentioned the use of the wild card search. It is perhaps surprising that 71.91% of LLB students used the wild cards while 54.16% used truncation.
Search and Retrieval method used in Legal Information System
Note: n = 92
Method of Learning
In order to more fully understand the methods of using e-resources, users were provided with five options; i.e. workshops, tutorials, a printed manual, one-to-one tuition or any other method. The table below shows that 48.91% of the respondents learnt by their own, while 32.60% learnt with the help of friends or colleagues; 13.04% of the respondents stated they learned with the help of library staff. Only 5.43% of respondents stated they learnt the use of e-resources by trial and error.
Method of using e-Resources LLB and LLM students
Note: n = 92
Rating the Quality of Legal Information Database
The survey listed ten types of e-resources and respondents were asked to rate the quality as available through the library portal. Four options were given: excellent, very good, good and poor. Table 6 shows that the majority of respondents agreed that the quality of resources are excellent.
Rating the Quality of Information Database by LLB and LLM students
Note: Respondents were asked to tick multiple options in the questionnaire.
Training for Using of Legal Information Databases
The data revealed that the majority of law faculty members – i.e. 51.68% – needed training to use journal portals, while 48.38% replied that they did not require any training to use e-resources. 61.22% of LLB students responded that they needed training, while 51.61% of LLM students requested training in order to learn the necessary skills to use legal information systems. When asked to respond about the various methods of training, Table 7b indicates that 43.75% stated that the tutorial method should be the adopted approach for legal information database training whilst 14.53% thought that the lecture method was preferable. 41.66% responded that the workshop method should be used.
Training to use Legal Information Database by LLB and LLM students
Note: n = 93
Mode of Training of using Legal Information Database
Note: n = 48
FINDINGS AND CONCLUSION
In the digital age, documents pertaining to law, available in electronic form, are critical to legal research. Electronic retrieval of legal material facilitates the process of legal objective research. Law libraries have a difficult task to satisfy their stake holders in terms of legal research with online legal information systems. The case study conducted in University of Delhi found that many students, particularly the LLB students, are aware of the range of electronic legal information services in the library but that many are not full conversant with their usage. The study also reveals that 33.70% of the respondents use these online databases for case law searching, while only 21.34 for research and development activities. Only 11.23% of respondents stated they used these databases for their studies and for updating their legal research. As far as frequency of usage was concerned, only 48% used these legal databases. 26.02% of the respondents used these resources occasionally while only 16.30% stated they used them sometime. As far as search and retrieval in database searching was concerned, most of the users preferred the simple search method and the boolean operators. Only expert users preferred the wild card and other proximity operators. Nearly half of the respondents revealed that they used these resources by ‘own learning’ and nearly 32.60% stated that they learned with the help of friends. A large number of respondents desired to have appropriate training in order to learn the use of the databases. It is clear that library and information professionals in India, as everywhere, play a major role in increasing the usage electronic services to expedite the process of legal research.
