Multi-Agent Research Tool (MART) Second Phase Madhukar Kumar.

Multi-Agent Research Tool (MART) Second Phase

Madhukar Kumar

Project Plan Project statement Multi-agent system that helps a

researcher/author in writing an article by automatically searching and dynamically generating links while an article is being written.

Action Items from the last presentation. Be able to filter search by websites, document format and

date. Formal requirement specification.

Updated Project Plan New cost estimate Effort = 3.2*(2.500)1.05 = 8.4 person-months

(earlier it was 4.9 months) Earlier the estimated size was approx 1500 LOC. Time = 2.5*(8.4)0.38 = 7.9 months (earlier it was

4.6 months) Productivity = 2500/7.9 = 316 LOC-month (earlier it

was 326 LOC-month) Staff = 8.4/7.9 = 1 person (Earlier 1 person)

Updated Project Plan New Gannt Chart

Architecture Design- Goals

Architecture Design - Role Model

Architecture Design – Agent Template

Architecture Design – Sequence of Events

Architecture Design Object model from code. Object model redesigned.

Architecture Design – Big picture

Architecture Design - Big picture

Architecture Design – Search Algorithm Read keywords and user preferences. If

keywords and user preferences have not been defined by the user then read the typed article and extract words.

Filter out words like “if”, “a”, “the” etc and create objects (KeyObject) for all other words. Store the frequency of use, ranking and actual word in the objects.

Architecture Design – Search Algorithm Search with following schema.

One AgentGoogle conducts a search with all the keywords (pick top 3 from results)

One AgentTeoma conducts a search with all the keywords (pick top 3 from results)

One AgentGoogle with top three ranked words (pick top 5 from results)

One AgentGoogle with second three (pick top 4 from results)

One AgentGoogle with next two (pick top 3 from results) One AgentGoogle with last two (pick top 2 from results)

Formal Requirement Specification – USE Model

Formal Requirement Specification – USE model – System Constraints -- Some of the keywords are also present in the RawResult context km: AgentManager inv keysalsoinresult: (km.keys.name->asSet) ->

intersection(km.resultfile.resultword.word->asSet)->notEmpty

-- If a word appears more number of times in the user inputted article -- its ranking will be higher (1 is higher than 2) than a word which -- appears lesser number of times in the article context key : AgentManager inv rankings : key.keys->forAll( r,s : KeyObject | r.nooftime >

s.nooftime implies r.rank < s.rank)

Formal Technical Specification – USE Model (Constraint Violation)

Formal Technical Inspection Checklist

Conclusion Demonstration of prototype Questions.