About

I am was a final year PhD student in Media futures group at UCL. At present, I work at VerizonMedia, New York as a Research Scientist. At UCL, I worked with Dr Emine Yilmaz on understanding search effort on different devices. I have also worked on search task extraction both for my thesis and at Google. PhD came after a year's of work at Naukri.com where I was responsible for extracting and structuring information from Resumes and Job sites. Prior to that, I was in IIIT Hyderabad, where I worked on Patent Information Retrieval (specifically query construction for patents) under the guidance of Prof Vasudeva Varma. I did bachelors in Computer Science from JIIT Noida.

I recently gave an invited talk on mobile specific IR at ECIR workshop Proactive IR [Pdf].

I helped in organizing Tasks Track at TREC in 2015-2016. I am a student organizor of a new and exciting track: TREC Complex Answer Retrieval (CAR, poster) along with Laura in 2017. Consider participating!

Publications

LIRME: Locally Interpretable Ranking Model Explanation
Manisha Verma, Debasis Ganguly
SIGIR 2019 [website]

Study of Relevance and Effort across Devices
Manisha Verma, Emine Yilmaz, Nick Craswell
CHIIR 2018 [website]

On utility of temporal embeddings for skill matching
Manisha Verma, Nathan Francis
DSHCM ICDM 2017 [website]

Search Costs vs User Satisfaction on Mobile
Manisha Verma, Emine Yilmaz
ECIR 2017 (short) [Pdf] [Poster]

Going Beyond Relevance: Incorporating Effort in IR
Manisha Verma
SIGIR DC 2016, [link]

Characterizing Relevance on Mobiles and Desktop
Manisha Verma, Emine Yilmaz
ECIR 2016, [link]

Category based Task Extraction
Manisha Verma, Emine Yilmaz
CHIIR 2016 (short), [Poster]

On obtaining effort based judgments for Information retrieval
Manisha Verma, Emine Yilmaz, Nick Craswell
WSDM 2016, [Pdf] [Poster] [Ppt]

Entity oriented Task Extraction from Query Logs
Manisha Verma, Emine Yilmaz
CIKM 2014 (short), [Pdf]

Bringing Head Closer to the Tail with Entity Linking
Manisha Verma, Diego Ceccarelli
ESAIR 2014 (workshop), [Pdf]

Relevance and Effort: An Analysis of Document Utility
Emine Yilmaz, Manisha Verma, Filip Radlinski, Nick Craswell, Peter Bailey
CIKM 2014, [Pdf]

Applying key phrase extraction to aid invalidity search
Manisha Verma, Vasudeva Varma
ICAIL 2011, [Ppt], [Pdf]

Exploring Keyphrase Extraction and IPC Classification Vectors for Prior Art Search
Manisha Verma, Vasudeva Varma
CLEF-IP 2011, [Pdf]

Patent search using IPC classification vectors
Manisha Verma, Vasudeva Varma
PAIR, CIKM 2011 (workshop), [Ppt], [Pdf]

Work Experience

PhD Software Intern (Google)
Was responsible for developing system to extract commercial user search tasks for ad targeting. It involved evaluating and testing several classification models with user, query and session oriented features.

Senior Software Engineer (Naukri.com)
We used Apache UIMA to build a resume parser. We also built job attributes extraction system that extracted fields such as job title, description, location etc from crawled webpages.

Research Intern (Microsoft AdCenter)
I worked on developing models for keyword extraction for sponsored search. We applied IR techniques and supervised approaches that combined ad data, query and click logs to extract phrases related to an advertisement.

Research Assistant (Search and Information Extraction Lab,IIIT-H)
We developed a customized cross-lingual search engine for Patents in English, German and French. Query expansion techniques were explored to improve retrieval performance. Assisted Btech students to construct cross-lingual dictionary for patents using Wikipedia.

Research Assistant (Center for Education Technology and Learning Science,IIIT-H)
We developed a customizable learning environment from existing tools (Trac, SVN, reST, SQLite). The system allows individual and group annotations, versioning of the students work, custom querying, and a uniform markup language to store content. Poster