Services - 73980-2018

17/02/2018    S34    - - Services - Prior information notice without call for competition - Not applicable 

United Kingdom-Exeter: Software package and information systems

2018/S 034-073980

Prior information notice

This notice is for prior information only

Services

Legal Basis:

Directive 2014/24/EU

Section I: Contracting authority

I.1)Name and addresses
University of Exeter
RC000653
Northcote House
Exeter
EX4 4QH
United Kingdom
Contact person: John Story
Telephone: +44 1392725430
E-mail: j.story@exeter.ac.uk
NUTS code: UKK4

Internet address(es):

Main address: http://www.exeter.ac.uk

Address of the buyer profile: https://uk.eu-supply.com/ctm/Company/CompanyInformation/Index/53042

I.2)Information about joint procurement
I.3)Communication
Additional information can be obtained from the abovementioned address
I.4)Type of the contracting authority
Body governed by public law
I.5)Main activity
Education

Section II: Object

II.1)Scope of the procurement
II.1.1)Title:

Text Analytics Software

Reference number: UOE/2018/005/JS
II.1.2)Main CPV code
48000000
II.1.3)Type of contract
Services
II.1.4)Short description:

VISTA AR is a European funded project led by the University of Exeter that is working closely with 8 partners including Exeter Cathedral in the UK and Fougères Castle in France. The focus of the work is to develop an understanding of visitor experiences. Specifically the project aims to develop this understanding using Text Analytics (TA) to analyse visitor feedback collected from cultural heritage locations.

II.1.5)Estimated total value
Value excluding VAT: 260 000.00 GBP
II.1.6)Information about lots
This contract is divided into lots: no
II.2)Description
II.2.1)Title:
II.2.2)Additional CPV code(s)
48461000
72212461
72212783
II.2.3)Place of performance
NUTS code: UKK4
Main site or place of performance:

Exeter.

II.2.4)Description of the procurement:

VISTA AR is a European funded project led by the University of Exeter that is working closely with 8 partners including Exeter Cathedral in the UK and Fougères Castle in France. The focus of the work is to develop an understanding of visitor experiences. Specifically the project aims to develop this understanding using Text Analytics (TA) to analyse visitor feedback collected from cultural heritage locations.

The use of advanced (TA) digital technologies will inform dimensions of experience quality within a Dashboard. These dimensions will subsequently be evaluated against visitor profiles (using demographic and psychographic data) provided by complimentary tools. The analysis of visitor experiences, as per the experience quality dimensions, will inform the design and development of devices to create new/enhance existing immersive visitor experiences. The project consortium is looking to engage with a software company that has a capability to develop a Software as a Service (SaaS) system that encapsulates advanced Natural Language Processing (NLP) algorithms to perform TA on visitor feedback data. The Text Analytics system will be developed to analyse both English and French visitor feedback. The TA software will reside on a cloud-based system at one of our partner sites and will be easily accessible and adaptable to the needs of a specific heritage location. The TA system will communicate with other tools (such as the Digital Profiling Tool) on the cloud. These tools will provide a corpus of textual input for subsequent analysis and visualisation by the TA system. The principal output of the system will be management reports for cultural heritage sites, on a periodic basis, that present experience quality dimensions and overall experiential outcomes.

The visitor intelligence provided by the TA system (and other associated tools) will be used to continuously improve the value propositions made to different visitor groups and to inform service innovation.

All responses to this PIN to be received via CTM no later than 15.3.2018 Midday.

Suppliers that do not respond to this PIN will be able to express their interest in participating in any procurement opportunity that may result from the market research, by responding to any Contract Notice. Suppliers will suffer no prejudice for not participating in the university's market engagement. All tender activity will be conducted via CTM (University's e-Tendering portal), including all tender related correspondence.

II.2.14)Additional information
II.3)Estimated date of publication of contract notice:
02/07/2018

Section IV: Procedure

IV.1)Description
IV.1.8)Information about the Government Procurement Agreement (GPA)
The procurement is covered by the Government Procurement Agreement: yes

Section VI: Complementary information

VI.3)Additional information:

The University of Exeter is looking to the market to provide a solution to meet our requirements for a Natural Language Processing system to analyse visitor feedback. The University wishes to explore the availability of solutions that are free of ongoing commercial costs such as those charged for the use of commercially available APIs.

Textual data will be continually extracted from third party and bespoke platforms (in English and French) and made available on the aforementioned cloud. The TA system would then be triggered by the Digital Profiling Tool (resident on the cloud) every time a unique heritage location will request for visitor intelligence reports. Textual data collected for the heritage location will be provided as input to the TA system. The goal is to understand and improve visitor experience for cultural heritage sites from the analysis of textual data. In particular, the system should be able to:

Identify entities/concepts from the input text using approaches including (but not limited to) Hand-Crafted Rule-based Concept Recognition (CR), Machine Learning-based CR or Hybrid CR. Basic (pre-processing) steps for concept identification would include:

— Text Tokenization and Segmentation,

— Stop Word Removal,

— Lexical Analysis (Part of Speech Tagging),

— Morphological Analysis (Concept Root Derivation and Synonym Detection), and

— Syntactic Analysis (Word Association Analysis).

Classify the identified concepts into domain-specific concept types (i.e., types corresponding to a visitor’s experience at a heritage location). Preliminary (domain-specific) concept types and libraries will be provided by the University of Exeter.

Concept classification can be performed using Classification algorithms including (but not limited to) Fuzzy Logic, Support Vector Machines, Neural Networks, etc.

Perform syntactic analysis (Word Association Analysis) and semantic analysis on these concepts using approaches such as (but not limited to) Latent Semantic Analysis. Such analysis would be beneficial for dealing with ambiguous concepts as well as analysing concept influence on experience quality dimensions.

Perform a statistical opinion analysis based on the aforementioned investigation conducted on identified and classified concepts. In particular, the system should statistically inform various Dashboard experience quality dimensions concerned with a visitor’s experience at a unique heritage location (such as staff knowledge) in terms of positive or negative visitor feedback. Management reports based on the analysis will, thereafter, be produced as per a heritage location’s requirement (i.e., at fixed periods of time or for a customized time interval as specified by the heritage location). The system should also deploy functionalities for comparative analysis among management reports of different time periods if and when specified by a heritage location. These reports would be made available on the cloud to be used for further processing by other tools to generate an integrated visitor intelligence.

The University of Exeter would like to hear from suppliers willing to share information about their general product offering and approach to enable us to:

— gauge the likely level of interest in the project from the market,

— better understand what solutions and services are available in the market,

— explore how the options could be supplied, and

— consider the likely costs.

Responses in the form of PDF documentation and/or prerecorded webinars/videos etc. are anticipated and requested at this stage.

Responses in pdf format and/or prerecorded webinars/videos etc. are requested at this stage and information must be submitted via CTM using the following URL: https://uk.eu-supply.com/login.asp?B=EXETER suppliers are required to search for and select the appropriate tender from the list of current opportunities.

VI.5)Date of dispatch of this notice:
16/02/2018