1 / 15

Genetic Algorithms for Topical Web Search: A Study of Different Mutation Rates

Genetic Algorithms for Topical Web Search: A Study of Different Mutation Rates . Roc ́ ıo L. Cecchini †, Carlos M. Lorenzetti‡, Ana G. Maguitman ‡ and N ́ elida Beatriz Brignole †§ . Agenda. INTRODUCTION GENETIC ALGORITHMS GENETIC ALGORITHMS FOR EXPLORING QUERY SPACE

aman
Télécharger la présentation

Genetic Algorithms for Topical Web Search: A Study of Different Mutation Rates

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Genetic Algorithms for Topical Web Search: A Study of Different Mutation Rates Roc ́ıo L. Cecchini†, Carlos M. Lorenzetti‡, Ana G. Maguitman‡ and N ́elida Beatriz Brignole†§

  2. Agenda • INTRODUCTION • GENETIC ALGORITHMS • GENETIC ALGORITHMS FOR EXPLORING QUERY SPACE • SYSTEMARCHITECTURE • THE EFFECT OF DIFFERENT MUTATION RATES • CONCLUDING REMARKS

  3. INTRODUCTION • Harvesting topical content is a process that can be done by formulating topic-relevant queries and submitting them to a search engine. • In this scenario, selecting good query terms can be seen as an optimization problem where the objective function to be optimized is based on the effectiveness of a query to retrieve relevant material.

  4. INTRODUCTION • Solving this optimization problem is challenging because the query space has a huge number of dimensions. • TouseGAtosolveit,becausefollowingcharacteristics: • 1.It is usually sufficient to identify suboptimal solutions • 2. we may be interested in finding many high-quality queries rather than a single one.

  5. GENETIC ALGORITHMS • Selection:Parents are selected to produce offspring, favoring those parents with highest values of the fitness function. • Crossover:Crossover of population members takes place by exchanging subparts of the parent chromosomes • Mutation:mutation is the result of a random perturbation of the chromosome (e.g., replacing a gene by another)

  6. GENETIC ALGORITHMS • Step 1: Start with a randomly generated population • Step 2: Evaluate the fitness of each individual in the population • Step 3: Select individuals to reproduce based on their fitness • Step 4: Apply crossover with probability Pc • Step 5: Apply mutation with probability Pm • Step 6: Replace the population by the new generation of individuals Step • 7: Go to step 2

  7. Selection-fitnessfunction

  8. Selection

  9. Crossover • The crossover operator used in our proposal is known as single-point. • new queries: • 1.the first n terms are contributed by one parent • 2.the remaining terms by the second parent, where the crossover point n is chosen at random.

  10. Mutation • These changes consist in replacing a randomly selected query term t1by another term t2. The term t2is obtained from a mutation pool . • Mutation pool:is updated with new terms from the snippets returned by the search engine

  11. SYSTEMARCHITECTURE

  12. THE EFFECT OF DIFFERENT MUTATION RATES

  13. Evaluationofqueryquality

  14. Conclusion • 1. The techniques presented in this paper are applicable to any domain for which it is possible to generate term-based characterizations of a topic. • 2. higher mutations rates induce a more thorough exploration of the search space • 3. As part of our future work we expect to continue testing different settings for the GA parameters (population size, crossover probability, mutation probability) as well as other selection methods.

  15. Q&A

More Related